Understanding Google AdWords therefore requires asking what was new at each stage. Paid search placements existed before Google. So did click-based advertising. The system’s significance developed through a combination of search, a purchasing interface, allocation rules and feedback to advertisers. Numerous decisions about terms, messages and expenditure became parts of a continuously editable market.

For marketing history, this was more than a new advertising format. Part of media buying moved into software operated by the buyer. A business had to translate its intentions into settings, examine results and decide again. AdWords consequently matters as a change in working practices. Planning, publication and observation moved closer together, while uncertainty about advertising effectiveness remained.

23 October 2000: Direct Access to Advertising Purchases

On 23 October 2000, Google announced the immediate availability of AdWords. Approximately 350 businesses and agencies had participated during the beta earlier that month, according to the company. Advertisers could establish and monitor campaigns directly through Google’s website (Google, 2000).

The participation figure is a supplier’s statement about an early customer group. It should not be read as a description of an already mature global advertising market. The more revealing feature is the relationship proposed between the service and its customers. Google sold an advertising opportunity while supplying tools through which buyers could perform some of the work themselves.

The interface acquired an organizational role. It guided decisions that might otherwise require individual exchanges with a sales representative. Selecting terms, supplying copy, determining expenditure and examining performance became connected activities within one environment. The software served both as an ordering route and as a workplace.

Easier access did not necessarily make the underlying decisions simple. A quick technical start could not establish which terms were economically appropriate or which message would persuade. This tension was central to the offer: entry could become easier while responsibility for many consequential choices remained with the advertiser.

The Original Model: Impression Pricing Rather Than Pay-per-Click

Christopher Saunders’s contemporary report described prices of $10 to $15 per thousand impressions and a minimum purchase of $50. It identified text advertisements to the right of search results and distinguished the service from Google’s existing premium program (Saunders, 2000). The initial billing event was therefore the delivery of an impression.

CPM, cost per mille, specifies a price for a thousand impressions. CPC, cost per click, specifies payment for a click. These units describe different events. An advertisement can appear without attracting a click, and a click can occur without producing a purchase. The billing choice determines which event triggers a charge.

The distinction matters to the relationship between buyer and supplier. Impression pricing initially purchases opportunities to be noticed; subsequent visits have to be examined separately. Click pricing ties payment to a further response. Even then, the advertiser remains exposed to the possibility that the visit produces little economic value.

The early price schedule should not be confused with later auction bids. A listed price and a competitive allocation mechanism organize buying differently. AdWords in 2000 was an important step towards self-service purchasing. The later combination of click pricing, competition and advertising response belongs to the subsequent development.

Earlier Precedents: Neither Paid Search Nor Click Billing Began at Google

Daniel Fain and Jan Pedersen’s historical survey identifies Open Text’s short-lived Preferred Listings in 1996. It also describes a click-based agreement between Yahoo and Procter & Gamble that year and identifies LinkStar as an early provider of CPC pricing for search advertising (Fain and Pedersen, 2006).

These examples concern different components of the eventual system. Preferred placement concerns an entry’s position. Click billing concerns the payment unit. A purchasing interface concerns advertiser access. None establishes, by itself, that the complete later arrangement already existed. The history becomes clearer when the components are distinguished before their combination is explained.

The distinction is particularly valuable for an international history of marketing. A later market leader can overshadow earlier approaches in public memory. Familiarity with a brand should not determine which contributions remain visible. A brief experiment can identify an important commercial question even when another organization subsequently finds a more durable implementation.

AdWords therefore belongs within a longer commercialization of search. Its significance cannot be adequately described as inventing the entire field. What requires explanation is how one service developed within an emerging market and why its combination of access, distribution and feedback became consequential.

GoTo and Overture: An Established Alternative

Overture’s annual report for 2002 dates its official service launch to 1 June 1998 and its change of name from GoTo.com to 8 October 2001. It describes bid-ranked placements charged by the click. At the end of 2002, it reported approximately 80,000 advertisers and operations in the United States, United Kingdom, Germany, France and Japan (Overture Services, 2003).

Google was entering a market with alternative intermediaries, rules, distribution arrangements and accumulated experience. Advertisers were not deciding only whether search advertising was possible. They also had to assess the audiences available through different services, the prices and allocation rules, and the work involved in maintaining a campaign.

Overture’s international presence also prevents a purely American account of market expansion. Search advertising was organized across languages and economies. A technical arrangement could travel while the available offers, customer needs and vocabulary differed. A common platform did not establish identical conditions in every market.

The comparison cautions against letting the eventual outcome explain the entire past. Google’s later importance does not remove the need to examine contemporary competition. One business can establish important components earlier while another combines them differently. Innovation and enduring dominance remain separate matters for historical explanation.

What Self-Service Changed

Saunders reported on the launch day that advertisers could create and edit messages and obtain statistics without going through a media representative (Saunders, 2000). The organizational change was direct access to campaign operations. Each alteration no longer necessarily required a new instruction to a personal sales contact.

Work was redistributed. Some activities associated with an individually handled sale became standardized inputs. This could make smaller purchases easier to accommodate because every decision did not require a separate negotiation. For the supplier, one technical environment could support interactions with many buyers.

Customers also moved closer to the advertising process. They could change settings but needed to understand them. An accessible control does not establish which adjustment would be useful. Access to a tool and the ability to use it well are different requirements.

Calling this a democratization of advertising captures only part of the development. Access could broaden while knowledge, available time and financial capacity remained unevenly distributed. The more precise historical change was a standardized route into campaign purchasing and editing. It did not make all participants’ competitive prospects equal.

This also changes how the interface should be interpreted. Its significance was not limited to convenience. By deciding which inputs buyers supplied and which outcomes they could inspect, the system shaped the form in which advertising decisions were made. The purchasing environment became part of the practice itself.

Keywords Became Decisions About the Market

The launch announcement already listed negative keywords and phrase and exact matching. Google’s concrete example connected an eyeondomain.com domain advertisement with a query about buying a domain (Google, 2000). Advertisers were specifying conditions for appearance as well as supplying a message.

Choosing a term makes an implicit statement about the need an offer can satisfy. It requires translation between a company’s vocabulary and the language of potential customers. An internal product name may differ from the way someone describes a problem. A broad expression can also encompass several incompatible purposes.

Excluding terms is therefore part of defining a market. Attention is not automatically useful merely because it can be acquired. A supplier has reasons to draw boundaries when an offer is unsuitable for particular expectations, applications or places. Technical targeting makes such distinctions possible; judging their commercial meaning still requires knowledge of the offer and its customers.

The historical change was the translation of organizational assumptions into editable rules. Technology did not eliminate market understanding. That understanding had to be expressed in a form the system could process. Search advertising connected established questions about customers and benefits with specific choices of words and conditions.

International Reach Was Not Yet an International Campaign

In February 2001, Google reported search services in 26 languages and relationships with portals and websites in Europe and Asia during the previous year. It identified German, Japanese and French as particularly prominent among non-English queries at year-end (Google, 2001). These statements concern search usage, rather than an equivalent number of localized AdWords interfaces.

The distinction between an audience that can be reached and a market that can be served is fundamental. A supplier can technically attract attention abroad without being able to deliver an appropriate service there. Language, payment, fulfilment and customer support affect whether a contact is commercially meaningful. The advertisement begins a relationship rather than completing it.

Translation alone cannot settle the matter. A linguistically correct description may differ from local search vocabulary or create expectations the supplier cannot satisfy. International advertising therefore connects language work with knowledge of a particular market.

The historical development was a shared technical infrastructure that could provide access to different search contexts. It did not abolish their differences. Google’s and Overture’s international expansion should be understood both as the spread of a system and as the multiplication of concrete decisions facing its customers.

AdWords Select in 2002: Introducing Click-Based Pricing

On 20 February 2002, Google announced AdWords Select with CPC pricing. Ad ordering would combine the amount offered per click with click-through rate. Premium advertisements would retain impression pricing. Google also announced a Discounter intended to adjust the prices advertisers actually paid (Google, 2002).

Two changes need to be distinguished. A user response became the billing event, and a measure of response also entered the ordering of advertisements. An amount offered was therefore no longer a complete explanation of placement. Willingness to pay and likely response became connected within the allocation process.

This produced a different relationship between copywriting and purchasing. An advertisement had to attract attention while its supplier expressed an economically acceptable willingness to pay. Buying could no longer be understood entirely as choosing a predetermined space. The effectiveness of a message could interact with the terms of its distribution.

The announcement records the product offered in 2002. It cannot establish the complete set of ranking inputs used by every later system. The historical turning point is specific enough: click billing and response-sensitive allocation were brought into a self-service environment. That was a different stage from the October 2000 offer.

Why an Advertising Auction Is More Than a Price Comparison

Benjamin Edelman, Michael Ostrovsky and Michael Schwarz examined generalized second-price auctions for search advertising in 2007. They showed that their strategic properties differ from those of the Vickrey-Clarke-Groves mechanism. The term second-price therefore does not provide a general guarantee that reporting one’s true value is the best strategy (Edelman, Ostrovsky and Schwarz, 2007).

For marketing history, the important issue is market design. A platform has to establish how several positions are allocated, which information enters the ordering and how payments result. Those rules can influence participants’ conduct. An auction does more than uncover an already existing price: it supplies conditions under which prices are formed.

A higher position can offer more visits and entail additional expenditure. It cannot be valued independently of the advertiser’s economics. The supplier needs to consider what a further visit might contribute, as well as what competitors appear prepared to pay.

The research also explains why a simple highest-bid account is insufficient. Several positions, different response probabilities and competing objectives create a distinct allocation problem. Media purchasing became a subject for economic analysis of the mechanism itself, rather than only a comparison of published rates.

Click-Through Rate: A Response Signal With Economic Consequences

Hal Varian’s 2009 analysis explained the relationship between bids, expected click probability and search-engine revenue. Combining a bid with expected clicks connects allocation to expected revenue from an impression (Varian, 2009). The paper examines its contemporary auction setting, rather than every feature of later Google advertising systems.

The intuition is straightforward. A price per click generates revenue for the intermediary when a click occurs. Advertisements with the same click price but different probabilities of attracting clicks can therefore have different expected revenue implications. The payment unit has to be considered together with the likelihood of the event.

A high click-through rate does not establish high commercial quality after arrival. Copy can arouse curiosity and then disappoint. A precise message might attract fewer people while matching their purposes more closely. The advertiser’s ultimate objective and the response visible on the results page overlap without becoming identical.

This produced a more demanding relationship between creative work and measurement. Copy was neither decoration nor an isolated sales claim. Its response could be observed and connected to distribution. At the same time, the business had to avoid making one observable reaction its complete definition of success.

Conversion Tracking: Looking Beyond the Click

In a 2016 retrospective, Google product managers Tal Akabas and Frederik Kemner dated the introduction of AdWords conversion tracking to 2001. The early measure basically recorded whether a desired website action followed an advertising click (Akabas and Kemner, 2016). Measurement development should therefore not be merged indiscriminately with the later introduction of CPC pricing.

The sequence demonstrates that observation and billing can develop separately. A business can examine actions following visits while paying for impressions. It can also pay for clicks without adequately understanding what follows. Tracking and the payment model address different problems.

A conversion is initially a specified event. An order, registration and enquiry each mean something different. The organization must establish why it regards that action as significant. Particularly in a longer decision process, an initial signal does not constitute an economically completed transaction.

The connection nevertheless mattered for performance marketing. A campaign could be examined through a path from message to defined response. Additional data required additional precision: what was counted, how it was attributed and which economic conclusion could reasonably follow from it?

Click Prices and Customer Value Belonged to Different Calculations

Varian’s economic treatment connects the value of additional clicks with decisions about bids and positions (Varian, 2009). For the advertiser, it highlights a distinction: the amount charged for a click is an expenditure, while the contribution of the next visit is initially an expectation.

The reasoning does not require invented historical campaign figures. If only some visitors become customers, the value of a transaction must be considered across several visits. A lower proportion of successful contacts can make the same click price less attractive. A more valuable completed transaction can change the expenditure an advertiser can justify. Revenue alone is not necessarily the relevant value; costs and the remaining contribution also matter.

Potential reach does not answer this question. Nor does the total available budget determine what one further contact is worth. A budget limits an aggregate amount, whereas valuing a click concerns the relationship between additional expenditure and an expected outcome.

This helps explain why search advertising could become more closely connected to business calculation. Marketing had to bring together information about an offer, the likelihood of a transaction and its economic contribution. The platform supplied a purchasing mechanism. What an individual supplier could reasonably spend also depended on circumstances beyond that platform.

Later Research Complicated Simple Accounts of Effectiveness

Thomas Blake, Chris Nosko and Steven Tadelis conducted large-scale experiments at eBay to assess the incremental effects of paid search. Their 2015 paper found no measurable short-term benefits from the brand-keyword advertising examined in that setting. Effects for other keywords differed between new or infrequent users and frequent users (Blake, Nosko and Tadelis, 2015).

The findings are not a universal verdict on search advertisements. They concern the conditions, users and periods studied. Their historical importance lies in sharpening a question: would an observed purchase have happened without the advertisement? Attribution following a click and evidence of an additional effect answer different questions.

Search makes the distinction especially relevant because interest can precede exposure to the advertisement. Someone searching for a familiar company may already be close to buying from it. A subsequent purchase does not, by itself, establish how much of the behavior the advertisement caused.

AdWords history thus leads from easier observation to more demanding explanation. Additional numbers did not automatically make effectiveness unambiguous. They supplied more ways to describe a sequence. Evaluation still had to distinguish changes caused by advertising from existing demand and other influences.

From a Completed Order to Continuing Campaign Work

When settings and feedback sit within the same working environment, publication can be followed immediately by another decision. The campaign becomes less easily understood as a completed order. It becomes an object of repeated intervention.

Jansen’s preface explicitly emphasizes explaining differences in keyword and advertising performance, rather than concentrating on operating instructions alone (Jansen, 2011). This helps distinguish frequent alteration from learning. An editable campaign provides opportunities to act, but action does not automatically produce reliable knowledge.

Changing several settings simultaneously makes it difficult to attribute a subsequent difference to one intervention. Drawing conclusions from very few observations can overstate random fluctuations. Demand can also change over time. Evaluation must distinguish alterations to the campaign from alterations to its environment.

Organizations consequently faced a management task. They needed objectives, responsibility for changes and standards for judging results. Direct access could shorten administrative paths while still requiring coordination with the offer, sales and customer support. The campaign might generate a contact, but handling that contact involved work beyond the advertising system.

Self-service did not make expertise redundant. Its value could increasingly lie in selecting, interpreting and coordinating decisions. Standardized entry and specialist work could develop together rather than represent opposing outcomes.

Disclosure: Paid Visibility Needed to Be Recognizable

A 27 June 2002 FTC staff letter to Commercial Alert emphasized clear disclosure of paid placement and, in some circumstances, paid inclusion. It expressly withheld a final determination about the legality of the challenged practices (Federal Trade Commission, 2002). The document records a contemporary debate over recognizing commercially influenced selection.

The historical issue concerns the meaning of an interface. A results page brings different kinds of entries together, and users need to interpret what they are seeing. The relationship between advertisement and result is therefore not determined solely by background technical processes. Labels and arrangement are part of the intermediary’s work.

A relevant offer remains an advertisement when payment explains its access to attention. Identifying it as advertising does not establish that it is unhelpful. The distinction concerns the reason for its placement. That information helps the user understand the selection more accurately.

Trust was consequently connected to pricing. An advertising system needs people to continue using its interface and interpreting the offers presented there. The commercial value of intermediation depends partly on expectations about search. That relationship had to be maintained while new sources of revenue developed.

AdSense in 2003: Extending Beyond Google’s Search Page

On 18 June 2003, Google announced a self-service AdSense option for website publishers. It extended content-targeted advertising services introduced that March. Publishers could display AdWords text advertisements relevant to their pages and receive a share of click-generated revenue (Google, 2003).

A query and a page context are different signals. A query is an explicit user action. Page content initially describes the subject of a document. Either can supply information for advertising, but the circumstances differ. The extension was consequently more than an enlargement of the same search results page.

It also added roles. Alongside advertisers, searchers and the intermediary stood publishers providing further opportunities for exposure. Advertisers did not need a separate sales relationship with every participating website. Publishers could join an arrangement connecting demand for advertising with distributed space.

AdWords and AdSense describe different perspectives within that mediation. One addressed advertisers, the other publishers wishing to carry advertisements. Their connection shows how a purchasing system could become a broader infrastructure. The consequences reached the funding of content beyond Google’s original search interface.

Economic Scale and What Revenue Figures Can Prove

Google’s annual report for 2004 records advertising revenue of approximately $410.9 million in 2002, $1.421 billion in 2003 and $3.143 billion in 2004. It also records the termination of Premium Sponsorships on 1 January 2004 (Google, 2005). The figures include advertising on Google’s sites and its network; they do not isolate AdWords search revenue.

The figures establish rapid expansion of the advertising business, rather than profitability for every customer. Platform revenue and advertiser returns belong to different accounts. Aggregate growth also does not identify the contribution of an individual product alteration. More users, advertisers, distribution opportunities and changing behavior can operate together.

This prevents a familiar historical shortcut. A supplier’s success can make its service appear retrospectively successful for everyone involved. Outcomes can in fact differ. A useful allocation mechanism does not eliminate competition, uncertainty or mistakes by its participants.

The numbers therefore need to be read alongside the rules. Participation, allocation and payment remain central to explaining the milestone. Economic scale makes those institutional decisions consequential; it cannot replace analysis of them.

A Field of Knowledge and a Later Change of Name

Jim Jansen’s 2011 book Understanding Sponsored Search placed questions about keywords, advertising, prices and measurement at the center of its approach. Its preface distinguished underlying explanations from rapidly changing platform instructions (Jansen, 2011). The book documents the intellectual development of a field that could not be reduced to operating an account.

Campaign work combined precision in language, economic judgment and interpretation of evidence. A term had to fit a purpose, a message an expectation and an expenditure its possible benefit. Tools supported these decisions, but their relationship remained a matter for understanding rather than a sequence of mechanical inputs.

On 27 June 2018, Sridhar Ramaswamy announced the Google Ads brand, intended to represent a wider range of advertising capabilities than the original AdWords name (Ramaswamy, 2018). That later development should not conceal the specific form of the 2000 service.

AdWords’ historical contribution lies in the progressive combination of its components. Self-service changed access to purchasing. Click pricing changed the billable event. Auctions and response signals changed allocation. Tracking added opportunities for observation, while effectiveness research clarified the limits of inference. Together, these developments helped organize marketing as continuing work within a technical intermediary. The significance of 2000 is the beginning of that particular system’s development, whose later components must remain historically distinguishable.

References

Akabas, T. and Kemner, F. (2016) Simplifying conversion measurement: Saying goodbye to “Converted clicks”. Google, 25 July. https://blog.google/products-and-platforms/products/ads/simplifying-conversion-measurement/ (Accessed: 23 October 2026).

Blake, T., Nosko, C. and Tadelis, S. (2015) ‘Consumer heterogeneity and paid search effectiveness: A large-scale field experiment’, Econometrica, 83(1), pp. 155-174. https://doi.org/10.3982/ECTA12423. Accessible preprint: https://faculty.haas.berkeley.edu/stadelis/BNT_ECMA_rev.pdf.

Edelman, B., Ostrovsky, M. and Schwarz, M. (2007) ‘Internet advertising and the generalized second-price auction: Selling billions of dollars worth of keywords’, American Economic Review, 97(1), pp. 242-259. https://doi.org/10.1257/aer.97.1.242. Accessible preprint: https://web.stanford.edu/~ost/papers/gsp-final.pdf.

Fain, D.C. and Pedersen, J.O. (2006) ‘Sponsored search: A brief history’, Bulletin of the American Society for Information Science and Technology, 32(2), pp. 12-13. https://doi.org/10.1002/bult.1720320206.

Federal Trade Commission (2002) FTC staff closing letter: Commercial Alert response letter. 27 June. https://www.ftc.gov/sites/default/files/documents/closing_letters/commercial-alert-response-letter/commercialalertletter.pdf.

Google (2000) Google launches self-service advertising program. Press release, 23 October. https://googlepress.blogspot.com/2000/10/google-launches-self-service.html (Accessed: 23 October 2026).

Google (2001) Google achieves significant business, growth milestones in 2000. Press release, 1 February. https://googlepress.blogspot.com/2001/02/google-achieves-significant-business.html (Accessed: 23 October 2026).

Google (2002) Google introduces new pricing for popular self-service online advertising program. Press release, 20 February. https://googlepress.blogspot.com/2002/02/google-introduces-new-pricing-for.html (Accessed: 23 October 2026).

Google (2003) Google expands advertising monetization program for websites. Press release, 18 June. https://googlepress.blogspot.com/2003/06/google-expands-advertising-monetization.html (Accessed: 23 October 2026).

Google (2005) Form 10-K for the fiscal year ended December 31, 2004. US Securities and Exchange Commission. https://www.sec.gov/Archives/edgar/data/1288776/000119312505065298/d10k.htm.

Jansen, J. (2011) Understanding Sponsored Search: Core Elements of Keyword Advertising. Cambridge: Cambridge University Press. https://doi.org/10.1017/CBO9780511997686. Preface consulted: https://assets.cambridge.org/97811070/11977/frontmatter/9781107011977_frontmatter.pdf.

Overture Services (2003) Form 10-K for the fiscal year ended December 31, 2002. US Securities and Exchange Commission. https://www.sec.gov/Archives/edgar/data/1060439/000095014803000429/v88074e10vk.htm.

Ramaswamy, S. (2018) Introducing simpler brands and solutions for advertisers and publishers. Google, 27 June. https://blog.google/innovation-and-ai/technology/ads/new-advertising-brands/ (Accessed: 23 October 2026).

Saunders, C. (2000) ‘Google launches self-service ad program’, ClickZ, 23 October. https://clickz.com/google-launches-self-service-ad-program/76737/ (Accessed: 23 October 2026).

Varian, H.R. (2009) ‘Online ad auctions’, American Economic Review, 99(2), pp. 430-434. https://doi.org/10.1257/aer.99.2.430. Accessible preprint: https://www.aeaweb.org/conference/2009/retrieve.php?pdfid=254.

The year 1998 marks the public description of a Google prototype and the foundation of a company whose subsequent scale was far from inevitable. It does not mark the invention of the search engine or the beginning of every activity later called search engine optimization. The more precise turning point was a combination: hyperlink analysis, textual relevance and an architecture intended to operate over large collections became parts of a working search system. Understanding that combination is more useful than reducing Google’s history to a single celebrated formula.

For marketing history, the development illuminates a particular change in the conditions of visibility. Organizations continued to publish their own information, but being found also depended on relationships outside their websites and on selection by an intermediary. SEO can therefore be understood through a relationship between an organization’s documents, their place in the wider web and the purpose of a person searching for information. The relationship extends beyond the language of a promotional message.

Why 1998 Was a Milestone Rather Than a Beginning Without Precedents

Google’s history is often narrated backwards from its later prominence. A familiar brand makes the early product seem destined for success, and the product makes its underlying research appear inevitable. A more revealing approach starts with the contemporary evidence. What existed in 1998? Which problems did its creators identify? What could a prototype demonstrate, and what remained to be built?

Brin and Page’s 1998 paper, The Anatomy of a Large-Scale Hypertextual Web Search Engine, described a prototype with a full-text and hyperlink database covering at least 24 million pages. That figure describes the research system rather than complete coverage of the internet or an established position of market leadership (Brin and Page, 1998).

A research publication, a business foundation and a patent are different historical events. A publication exposes an approach to examination and discussion. A company provides an organization through which people can finance, maintain and market a service. A patent records a particular invention within a legal process. Collapsing the three into a single moment obscures the work needed to carry a promising method into regular use.

Google’s corporate account places Page and Brin’s meeting in 1995, identifies Backrub as an early project name and dates the company’s foundation to 1998. The garage belongs to the subsequent business story. It does not explain the entire research origin (Google, n.d.). As a company source, this account is useful evidence of how Google narrates its beginnings. The account becomes more informative when read alongside the research publications and institutional records.

Stanford, Digital Libraries and the Infrastructure Behind the Idea

The National Science Foundation traces part of Google’s research background to the Digital Library Initiative launched in 1994. Stanford’s project involved Hector Garcia-Molina and Terry Winograd. Page received support through that research environment, while Brin held an NSF Graduate Research Fellowship. Equipment also depended on institutional and industrial funding (National Science Foundation, 2004).

This background changes the interpretation of the later marketing phenomenon. Before Google became an important route to prospective customers, it was an effort to improve access to a growing and uncontrolled collection of information. The digital-library perspective brings collection, description, connection, storage and retrieval into the same frame. Marketing significance developed from that work, but it should not replace its original informational purpose.

A founder-centered story can underestimate what a research institution contributes. Technical facilities matter, but so do discussion, criticism, shared methods and procedures for moving knowledge into practical use. Such resources are not simply scenery behind an individual insight. They help determine whether an observation can become a testable method and whether that method can become a usable prototype.

For marketing historians, the connection is between research institutions and market infrastructure. Technologies that alter markets do not always begin as marketing tools. Solving an information problem can create an intermediary service; that service can subsequently support new practices, specialist occupations and commercial opportunities. Google in 1998 offers a clear example of that sequence.

Before Google: Web Search and Descriptive Links

Oliver McBryan’s World Wide Web Worm at the University of Colorado provides a concrete precursor. His paper for the first international WWW conference at CERN in Geneva in 1994 described searching through document titles, reference hypertext and components of URLs. It also presented GENVL, a hierarchical virtual-library system (McBryan, 1994).

Several approaches to discovering web resources therefore existed years before the Google paper. Descriptive links are particularly relevant: their use did not begin without precedent at Google. The distinction is between an available information signal and its subsequent combination with other techniques in a different system.

The conference setting also illustrates the international research environment of the early web. Work from Colorado was discussed at a European research center. Google’s subsequent corporate importance is easier to understand when that earlier technical work remains visible, rather than being compressed into an indistinct period before the market leader.

Earlier Global Precedents: Citation Analysis Before Web Search

The idea of discovering documents through their relationships with other documents long predated the web. In 1955, Eugene Garfield proposed a citation index for scientific literature. References could expose connections between works that a purely topical arrangement might miss. His subject was scientific documentation, rather than the ranking of commercial websites (Garfield, 1955).

The distinction between a topic and a relationship is fundamental. Two documents can share a term while addressing different problems. Other documents can use different terminology while contributing to the same discussion. A reference adds another kind of evidence: someone has connected one work to another. The connection need not replace textual description to make that description more useful.

Gabriel Pinski and Francis Narin developed a related but distinct contribution in 1976. Their method assessed the influence of scientific journals through citation relationships, weighting references according to the influence of their sources. They formulated the problem through a citation matrix and an eigenvalue approach (Pinski and Narin, 1976).

This was an important precedent for recursively weighted influence. It does not mean that a network of scientific journals and the open web are equivalent, or that a commercial search service had already been invented. A mathematical relationship can recur across fields while the practical problems differ considerably. The web brought questions of access, changing documents, heterogeneous publishing practices and operation at a different scale.

Massimo Franceschet’s historical review situated PageRank among related methods in bibliometrics, sociometry and econometrics. That perspective places the technique within an international intellectual history, rather than an exclusively Californian origin story (Franceschet, 2011).

The correction does not diminish Page and Brin’s achievement. It makes the achievement more specific. Historical importance can lie in a successful combination and implementation, rather than in creating every underlying idea without precedent. For a museum of marketing, that distinction is valuable: many influential market practices emerge through the transfer and recombination of knowledge developed elsewhere.

Contemporary Alternatives: Kleinberg’s Authorities and Hubs

Research into the structure of the web also developed along parallel paths. Jon Kleinberg’s Authoritative Sources in a Hyperlinked Environment appeared in the Journal of the ACM in 1999, following a 1997 research report and a 1998 conference version. Its HITS approach distinguished between authorities and hubs: important sources and pages that point towards such sources (Kleinberg, 1999).

An excellent directory and an authoritative explanation need not perform the same function. A directory can be valuable because it leads readers towards useful material. An explanation can be valuable because good directories identify it as worth consulting. The relationship illustrates how the web can contain several forms of informational importance rather than one undifferentiated kind of popularity.

HITS was not another name for PageRank, and it should not be presented as a component of the early Google system. Its relevance here is comparative. Different researchers were exploring how hyperlinks could support judgments beyond navigation. A method aimed at a topic-related collection and a method that calculates a more general importance measure do not ask exactly the same question.

For organizations, this comparison reveals the limits of treating every website as though it played the same role. A site might be an original source, a guide, a commercial presentation or a collection of references. A useful information environment can require all of these functions. The historical research was more differentiated than the later shorthand of simply accumulating backlinks suggests.

How PageRank Works: Weighted Relationships Rather Than Link Totals

In the classical PageRank model, a page’s importance depends on incoming links, the importance of their sources and those sources’ outgoing links. A random surfer follows links and occasionally restarts elsewhere; long-run visits provide an intuitive interpretation of the scores (Franceschet, 2011).

The method is recursive because the value assigned to one document depends on other documents whose values are themselves determined within the network. It is therefore different from an ordinary count in which each incoming reference contributes an identical unit. The result reflects a structure of relationships rather than a collection of independent votes with equal weight.

A conceptual example makes the distinction clear. Suppose two company pages each receive one link. One link comes from a page isolated from the wider network; the other comes from a page with a strong network position. A simple count records a tie. A recursive method can distinguish between the two situations. This is an illustration of the model, not a reconstruction of actual Google results or a prediction of particular search positions.

The resulting score does not directly establish whether a product is suitable, a price is fair or a claim is true. Those judgments require other evidence. The technique’s strength is to turn relationships into a computable measure. Its limitation is that the measure captures a specific form of importance, rather than every possible meaning of quality.

That distinction helps explain PageRank’s relevance to marketing without exaggerating it. A technical score can structure opportunities for attention. It cannot stand in for a complete assessment of an organization or its offer. Customers may have good reasons to trust a less visible supplier or reject a highly visible one.

PageRank and Search Relevance Answer Different Questions

A strongly connected page can still answer a particular question poorly. A relatively obscure document can contain exactly the information required. This separates a page’s general position within a link network from its relevance to an individual query. A single importance score cannot, by itself, determine a useful ordering for every possible search.

The prototype also used textual signals, including anchor text, word positions and proximity. Link descriptions could be associated with the destination page, adding information supplied outside that page itself (Brin and Page, 1998).

This introduces a revealing distinction for corporate communication. An organization describes itself, but others can describe it differently. A specialist resource might receive a precise descriptive link, while a general corporate presentation is referenced only by company name. Neither case establishes a predictable ranking outcome. They illustrate why external description and self-description are separate sources of information.

The marketing implication is that a website is not a completely enclosed communication space. Its public meaning can emerge partly through the ways other participants locate and characterize it. Public relations, specialist recognition and publication can therefore be examined alongside organic visibility, without assuming that every such activity was originally performed as an SEO tactic.

The relationship is particularly useful for distinguishing communication from control. An organization controls much of its own published language. It has less direct control over the interpretations and connections created by others. Search brings these different contributions into a common information environment. That makes visibility a product of participation in a wider system, rather than simply a consequence of the firm’s intended message.

From an Algorithm to a Search Service

A ranking method is not yet a dependable search product. Information has to be gathered, processed, stored and made available in response to queries. Search quality and large-scale implementation must therefore be considered together when examining search as a service.

For users, the connection is straightforward. An excellent answer has limited value if it cannot be retrieved when required. A fast response is disappointing if it does not address the request. The service depends on several activities functioning together. Concentrating exclusively on PageRank can conceal the chain of work that precedes the visible results page.

This is also important for marketing history. It connects the promise of discoverability with the practical availability of information. A persuasive corporate story cannot substitute for an accessible description of an offer. An extensive technical explanation may be less useful if a reader cannot locate the relevant detail. These are analytical implications of search as a service, rather than claims about the precise weighting of present-day ranking factors.

Organizations face a related problem internally. Knowledge may exist within a company without existing in a form that outsiders can find or understand. Publishing requires decisions about what to explain, how to identify it and where to place it. Search adds an intermediary between that published material and its prospective reader. The entire path matters to the outcome.

The achievement associated with 1998 therefore concerned both a way of evaluating documents and the work of incorporating evaluation into a practical system. A mathematical insight and an operational information service belong to related but distinguishable histories. Their connection is precisely what makes the early Google prototype significant.

Patent, University and Company: Different Forms of Attribution

US Patent 6,285,999, Method for node ranking in a linked database, names Lawrence Page as inventor and Stanford as original assignee. It records priority on 10 January 1997, an application on 9 January 1998 and a grant on 4 September 2001. These are separate procedural dates (Page, 2001).

There is no contradiction between the joint scientific authorship of Brin and Page and the inventor named in this particular patent. The documents serve different purposes. Research authorship, patent inventorship and company foundation should each be described using evidence appropriate to that form of attribution.

A contemporary account from Stanford’s technology-transfer office adds another part of the story. Rich Scholes reported in 2000 that the technology had been disclosed to the Office of Technology Licensing in 1996 and discussed with internet companies before Page and Brin chose to establish their own business (Scholes, 2000).

This makes the startup visible as a route for developing the technology, rather than the predetermined conclusion of the research. An invention might be licensed, developed within an existing organization or used as the basis for a new firm. These arrangements affect who operates the resulting service and how responsibility for its development is organized.

For business history, the lesson is methodological as well as substantive. The most famous name attached to a product is not always the correct answer to every question about ownership, authorship or invention. Precision does not require a cumbersome legal narrative. It requires keeping the roles distinct enough that one document is not asked to prove something it was never intended to establish.

What Organic Visibility Changed for Marketing

The enduring connection with SEO begins with a different task for organizations. It was no longer sufficient to consider only what the business wanted to say. A company also needed to consider which information a searcher required, how that information existed as documents and how those documents were situated within a larger information environment. Organic visibility brought publication, technical implementation and public recognition into a closer relationship.

An available offer is not the same thing as an answered question. A company may sell an appropriate product without explaining its suitability clearly. It may have a remarkable history while the visitor needs dimensions, application details or a dependable contact address. Looking at information through the intended reader’s purpose changes the priorities of corporate publication.

This should not be presented as a complete marketing doctrine already formulated in 1998. It is a historical interpretation of the development. As search becomes a more important intermediary, translating organizational knowledge into publicly intelligible information becomes more consequential. The work involves editorial judgment, subject knowledge, technical implementation and communication together.

The development did not make older forms of reputation irrelevant. Helpful explanations and specialist recognition were valuable before Google. What changed was their possible connection to an automated infrastructure of discovery. A document written for people could also leave relationships that a search system could process. Public usefulness and machine interpretation could interact without becoming identical.

For managers, this complicates the boundaries of marketing responsibility. A marketing department may shape a message, but information quality can depend on engineers, customer-service staff, product specialists and editors. Organic visibility exposes the interdependence of these contributions. It is less easily understood as the isolated output of an advertising campaign.

Search Intent: A Later Concept That Clarifies the History

Andrei Broder’s 2002 taxonomy distinguished informational, navigational and transactional web searches. These concerned acquiring information, reaching a particular site or undertaking an activity; transactional did not mean purchasing alone. The taxonomy appeared after the Google paper and should not be retroactively attributed to the 1998 prototype (Broder, 2002).

The distinction nevertheless helps interpret the marketing change. Someone entering a company name may want to reach a known destination. Someone asking a technical question may need an explanation. Someone intending to use a service needs a suitable path towards action. The same corporate page can serve these purposes unequally, even when its general prominence remains unchanged.

Search intent and PageRank are therefore not interchangeable concepts. One concerns the purpose of a request; the other concerns relationships between documents. Keeping them separate clarifies the difference between being discovered and providing a satisfactory response. The fact that a user reaches a page does not settle whether the page helps them accomplish their task.

Intent also remains an interpretation rather than a direct view into a person’s mind. A short query can be ambiguous, and the same phrase can support different purposes. Marketing history should preserve that uncertainty. Describing search as purposeful does not require pretending that the purpose can always be read perfectly from the words entered.

The Limits of the Index: The Web Exceeded Every Results List

In 1999, Steve Lawrence and C. Lee Giles reported unequal website coverage, delays in indexing new pages and limited coverage by individual search engines. In their contemporary investigation, no single engine indexed more than approximately 16 percent of the web under study. The figure belongs to that investigation, not to present-day coverage (Lawrence and Giles, 1999).

The finding makes an important distinction visible. A document may be absent from a collection, inadequately represented or placed low in the results for a particular query. From a company’s perspective, the immediate experience can be similar: the page is not found. Analytically, these are different situations requiring different explanations.

A search engine is consequently not a simple mirror of everything that has been published. It operates over an acquired collection and applies a particular ordering. Explaining the commercial prospects of a website solely through its position overlooks the prior question of inclusion. For newer or less connected publishers, publication and effective discoverability are especially important to distinguish.

The issue also gives the history an international dimension. A globally accessible network does not establish equal visibility for every information provider. Google’s development should not be turned into a story of immediate, uniform access to all the world’s web content. Its contribution has to be situated within the limits of the information environment in which it operated.

Manipulation and the Changing Economics of Links

Once visibility has commercial value, participants’ incentives can change. A link may be created to help a reader, but it may also be produced to influence an automated assessment. Evaluating a relationship can therefore alter the conditions under which future relationships are made. The information system and the behavior it measures become connected.

Monika Henzinger, Rajeev Motwani and Craig Silverstein discussed text spam, link spam and other quality problems in their 2003 paper Challenges in Web Search Engines. Their account shows that manipulation became an explicit research problem early. It does not establish that every subsequent response to manipulation was already present in 1998 (Henzinger, Motwani and Silverstein, 2003).

For marketing, the distinction is between creating something useful and creating a signal mainly because the signal is evaluated. The activities can resemble each other externally. A reference may help a reader pursue a subject; another reference may exist principally to affect a machine’s interpretation. Their outward similarity does not make their informational contribution identical.

The underlying problem extends beyond search. When a measured characteristic becomes an economic target, behavior can adapt around it. The measure may then contain a different mix of information than it did before. PageRank’s history offers a concrete setting in which to examine this interaction between evaluation and market conduct.

This is also why a purely mechanical account of SEO is historically incomplete. The meaning of a technical signal depends partly on the practices through which the signal is generated. Understanding the system requires attention to those practices, including the motives of publishers and the expectations of users. The network is both a technical structure and an outcome of human activity.

Importance, Truth and the Responsibility of Intermediaries

Citation indexing also reminds us that references can lead readers towards criticism and correction, rather than endorsement alone (Garfield, 1955). The same conceptual caution applies to web links. A page can be linked because it is disputed, documented or used as a counterexample. The presence of a connection does not fully reveal its communicative purpose.

Lucas Introna and Helen Nissenbaum argued in 2000 that search engines raise political and social questions as well as technical ones. Selection and ranking can give prominence to some websites and kinds of websites while excluding or marginalizing others. Their subject was the consequences of search mediation for the public information environment (Introna and Nissenbaum, 2000).

This perspective complements the company-centered account. Visibility is an opportunity for an individual supplier; across an information environment, it is also a distribution of opportunities. Material appearing prominently is more likely to enter a user’s field of attention. Other material can remain publicly available while becoming difficult for many people to encounter in practice.

That does not establish deliberate disadvantage in every instance of unequal visibility. It establishes the importance of examining mediation itself. A history of marketing should explain not only how firms seek attention, but also how the institutions carrying information organize routes towards that attention.

The distinction between importance and truth is particularly useful here. A system can provide a valuable way to order documents without certifying every claim within them. Users, publishers and researchers still have reasons to examine sources and context. Historical explanation becomes stronger when it neither dismisses the technical achievement nor assigns it powers beyond the problem it addresses.

From Research Performance to the Google Brand

Scholes’s 2000 account identified word of mouth as important to Google’s early growth and described the provision of search services to other companies. Google could therefore operate both as a destination and as a supplier of a search capability (Scholes, 2000).

The combination matters for brand history. A name can become familiar through a company’s own interface while the underlying service develops through use in other organizational settings. Perceived usefulness can arise from repeated successful encounters, rather than from promotional claims alone. The service experience becomes part of the explanation of recognition.

Financing nevertheless remains a separate question. The 1998 paper’s appendix discussed potential conflicts between advertising-funded search and users’ interests (Brin and Page, 1998). That discussion records an early concern about incentives; it cannot explain the later advertising business on its own.

Organic ranking, the sale of a search service and the sale of search advertising are connected but historically distinguishable subjects. Keeping them separate makes the 1998 contribution easier to assess. The prototype’s informational and technical achievement should remain visible before the account turns towards the later commercial arrangements surrounding search.

Page and Brin’s Contribution to Marketing History

PageRank made hyperlinks useful as more than routes from one document to another. Combined with other information, they could support an automated ordering of pages. Page and Brin’s distinctive contribution involved connecting that evaluation with a practically described search system. Earlier citation methods and contemporary alternatives belong within the explanation because they clarify both the available ideas and the new demands of the web.

For marketing, the relationship between publication and attention acquired an additional layer. The company’s own content continued to matter, but discoverability also depended on an ordering produced beyond the company’s website. Visibility had to be considered through documents, relationships, search purposes and technical mediation together. That is why the history reaches beyond the wording of advertisements.

The significance of 1998 is therefore precise. It did not settle every question of relevance or make business success mathematically predictable. It made publicly visible an influential way of using the web’s structure to improve search. For organizations, that development created another condition of being found: the importance of published information depended partly on how it was connected to information elsewhere.

References

Brin, S. and Page, L. (1998) ‘The anatomy of a large-scale hypertextual Web search engine’, Computer Networks and ISDN Systems, 30(1-7), pp. 107-117. https://doi.org/10.1016/S0169-7552(98)00110-X. Full text: https://research.google.com/pubs/archive/334.pdf.

Broder, A. (2002) ‘A taxonomy of web search’, ACM SIGIR Forum, 36(2), pp. 3-10. https://doi.org/10.1145/792550.792552. Full text: https://sigir.hosting.acm.org/files/forum/F2002/broder.pdf.

Franceschet, M. (2011) ‘PageRank: Standing on the shoulders of giants’, Communications of the ACM, 54(6), pp. 92-101. https://doi.org/10.1145/1953122.1953146. Accessible 2010 preprint: https://arxiv.org/abs/1002.2858.

Garfield, E. (1955) ‘Citation indexes for science: A new dimension in documentation through association of ideas’, Science, 122(3159), pp. 108-111. https://doi.org/10.1126/science.122.3159.108. Author’s archive: https://garfield.library.upenn.edu/papers/science_v122v3159p108y1955.html.

Google (n.d.) Our story: From the garage to the Googleplex. https://about.google/company-info/our-story/ (Accessed: 23 October 2026).

Henzinger, M.R., Motwani, R. and Silverstein, C. (2003) ‘Challenges in web search engines’, Proceedings of the Eighteenth International Joint Conference on Artificial Intelligence, pp. 1573-1579. https://www.ijcai.org/Proceedings/03/Papers/278.pdf.

Introna, L.D. and Nissenbaum, H. (2000) ‘Shaping the Web: Why the politics of search engines matters’, The Information Society, 16(3), pp. 169-185. https://doi.org/10.1080/01972240050133634.

Kleinberg, J.M. (1999) ‘Authoritative sources in a hyperlinked environment’, Journal of the ACM, 46(5), pp. 604-632. https://doi.org/10.1145/324133.324140. Full text: https://www.cs.cmu.edu/afs/cs/project/pscico-guyb/realworld/www/papers/kleinberg.pdf.

Lawrence, S. and Giles, C.L. (1999) ‘Accessibility of information on the web’, Nature, 400, pp. 107-109. https://doi.org/10.1038/21987.

McBryan, O.A. (1994) ‘GENVL and WWWW: Tools for taming the Web’, in Nierstrasz, O. (ed.) Proceedings of the First International World Wide Web Conference. CERN, Geneva, May. https://citeseerx.ist.psu.edu/document?doi=24a0993b3ac32a4a3bd15fd9aef408bcff6f31f1&repid=rep1&type=pdf.

National Science Foundation (2004) Google. 17 August. https://www.nsf.gov/impacts/google (Accessed: 23 October 2026).

Page, L. (2001) Method for node ranking in a linked database. US Patent 6,285,999 B1, granted 4 September. Original assignee: The Board of Trustees of the Leland Stanford Junior University. https://patents.google.com/patent/US6285999B1/en.

Pinski, G. and Narin, F. (1976) ‘Citation influence for journal aggregates of scientific publications: Theory, with application to the literature of physics’, Information Processing & Management, 12(5), pp. 297-312. https://doi.org/10.1016/0306-4573(76)90048-0.

Scholes, R. (2000) ‘Uniquely Google’, Stanford Technology Brainstorm, Spring. Stanford University, Office of Technology Licensing. https://web.stanford.edu/group/OTL/documents/0902_sp00.pdf.

There is a specific milestone associated with 1998. In their technical retrospective, Brent Smith and Greg Linden dated Amazon’s introduction of item-based collaborative filtering to that year. The well-known explanation by Greg Linden, Brent Smith and Jeremy York appeared in 2003. Deployment and publication are connected events, but they are not the same date (Smith and Linden, 2017; Linden, Smith and York, 2003).

This is a more precise history than saying that Amazon invented artificial intelligence or personalized recommendations in 1998. Earlier projects had already investigated how people’s experiences could help others select information. Amazon’s contribution lay in a consequential application to a large retail operation. It connected calculation and interface design with a commercial task: helping customers discover relevant products.

What 1998 meant, and what already existed

Amazon’s report for the 1997 financial year already referred to instant recommendations, collaborative filtering and other personalization features. It also described a gift centre introduced in November 1997 with editorial and personalized suggestions. Recommendations therefore did not begin at Amazon in 1998, although the report does not reveal every technical detail of the earlier arrangements (Amazon.com, 1998).

The distinction matters because personalization is a broad category. Greeting a returning visitor by name, changing displayed content and calculating a product selection are different activities. They do not necessarily depend on the same procedure. An editor’s selection is also different from a relationship inferred from behavioural data.

A careful timeline consequently treats 1998 as a milestone for a particular approach. It should neither absorb all earlier activity nor make that activity disappear. Company reports, technical publications and later recollections illuminate different parts of the development. Their dates cannot be substituted for one another without changing the claim.

Commercial improvement does not require inventing every visible feature again. An existing function can become faster, more suitable or more practical at scale. The historical advance may lie in organizing a familiar task more effectively. That is one reason the exact technical contribution matters more than a generic label of innovation.

The background included electronic documents, not just shopping

David Goldberg, David Nichols, Brian Oki and Douglas Terry published their Tapestry paper in 1992. Developed at Xerox’s Palo Alto Research Center, the system belongs to the early history of collaborative information filtering. Its setting involved electronic documents and human judgements rather than an Amazon shopping basket (Goldberg et al., 1992).

The origin is revealing. Recommendation does not have to begin with a transaction. It addresses the relationship between an abundance of alternatives and limited human attention. Other people’s experience can help an individual choose without examining every possibility. Retailing placed that selection problem within a commercial setting.

The distinction from straightforward text retrieval is also useful. A search can ask which documents contain particular terms. A recommendation can ask which material may fit earlier judgements or interests. The tasks can support one another without relying on identical signals.

The history should therefore include communication and information research alongside the famous commercial application. A successful retail system did not emerge from an intellectual vacuum. Marketing history becomes more explanatory when it identifies the earlier problems and research communities involved, rather than treating the best-known business as the origin of every underlying idea.

GroupLens and Ringo made preference relationships computational

The 1994 GroupLens paper described predicted preferences for Netnews articles based on readers’ ratings. Paul Resnick and his co-authors presented an open architecture in which rating servers gathered and distributed judgements. Earlier agreement was used to help anticipate later interests. This was a concrete application outside online retail (Resnick et al., 1994).

In 1995, Upendra Shardanand and Pattie Maes presented Ringo, a system recommending music artists and albums. Their paper discussed use by more than 2,000 people and comparisons among several algorithms. These different domains establish a research history before Amazon’s item-based milestone (Shardanand and Maes, 1995).

“Collaborative” did not mean that participants had to meet and compose a recommendation together. Their contributions could be combined within a procedure. The cooperation was organized through recorded judgements, and people did not need personal relationships with one another for the system to operate.

That produced a different kind of mediation. A suggestion could depend on people the recipient had never met. Rules were needed to determine which similarities mattered and how evidence should be combined. The software did not create human preferences from nothing. It changed how traces of those preferences could be aggregated and passed on.

Word of mouth became a reference point, not an exact duplicate

Paul Resnick and Hal Varian’s 1997 introduction situated recommender systems alongside the ordinary practice of consulting others when personal experience is insufficient. They favoured the broader term “recommender system” because this assistance need not always involve the same collaborative-filtering process (Resnick and Varian, 1997).

The analogy explains some of the attraction. Nobody can read every book or try every available product. Guidance can reduce the effort of choosing. Computational mediation nevertheless changes the basis of confidence. A friend’s advice comes through a recognizable relationship; an algorithmic suggestion may draw on many relationships the recipient cannot inspect.

This raises a marketing question: why should somebody regard the output as useful? Its statistical basis might be reasonable without being obvious to the visitor. The interface influences how the suggestion is understood, not merely whether it is visible.

An unfamiliar recommendation accompanied by an intelligible connection may be interpreted differently from the same item presented without context. Translating calculation into something a person can assess is consequently part of the system’s achievement. It is not simply decorative packaging placed around a completed mathematical result.

Different approaches answered different questions

Marko Balabanović and Yoav Shoham’s 1997 Fab paper combined content-based and collaborative recommendation. Its existence shows that early research was not confined to one approach. Information about the material itself and evidence from its users could be brought together in different ways (Balabanović and Shoham, 1997).

A simple illustration separates the possibilities. A content-based method might connect two books because they discuss a similar subject. A collaborative method might connect them through patterns of selection or appreciation. The relationships need not coincide. Two items can look different in vocabulary or category while appealing to overlapping interests.

A bestseller list answers another question. It says something about general popularity rather than necessarily identifying what is appropriate for a particular person. An editorial selection can follow an explicitly chosen theme. None of these arrangements is inherently useless; each has different informational requirements and potential applications.

Historical explanation should therefore go beyond calling a system intelligent. What information enters it? Which relationship is calculated? How does that relationship shape what is shown? These questions identify the actual change more clearly than the label alone.

The item-based approach shifted the comparison towards products

Badrul Sarwar, George Karypis, Joseph Konstan and John Riedl examined item-based techniques in 2001, including similarity measures and ways of generating recommendations. They addressed quality and computational performance together. The work illustrates a broader technical investigation of relationships between items rather than only relationships between people (Sarwar et al., 2001).

The basic difference can be explained without a formula. A user-based approach asks which other people resemble a particular person. An item-based approach first identifies suitable relationships among products. Individual suggestions can then be assembled from the items associated with that person’s interests or activity.

An explicitly invented example is a reader who has bought an introductory photography book. Relationships associated with that item might help identify another book about composition. The connection need not result from identical words in the titles. It could arise from observed patterns of interest. This illustrates a principle rather than reconstructing an Amazon dataset.

The shift matters commercially because part of the work can be prepared before a live visit. Calculation, freshness and quality become design variables. A theoretically interesting suggestion is less useful to a functioning shop if it arrives only after an impractical delay. The retailer needs a procedure that operates within the visitor’s activity.

Scale also meant being ready at the right moment

Smith and Linden’s 2017 account explained the separation between prepared item relationships and rapid selection during use. It connected the approach with scale, responsiveness and intelligible links between suggestions and prior interests. These are retrospective statements by involved specialists, not an independent audit of every earlier feature (Smith and Linden, 2017).

The architecture reveals an easily overlooked condition of marketing usefulness. The result must arrive within the time available for a decision. A person browsing a store cannot be assumed to wait indefinitely for a marginally better list. Timely assistance performs a different role from an elaborate analysis completed after the visit.

Scale has several dimensions. More customers, more products and more current interactions do not necessarily increase the same part of the workload. Improving one bottleneck does not make all processing free. The achievement concerns a better allocation of work, rather than the disappearance of computational costs.

Claims about an efficient online stage should therefore not be read as claims that preparation, storage and maintenance require no resources. A recommender is an operational arrangement with several stages. The short visible list is an output of that arrangement, not a complete picture of its cost or complexity.

A patent supplies another dated record

US Patent 6,266,649, “Collaborative recommendations using item-to-item similarity mappings,” names Gregory Linden, Jennifer Jacobi and Eric Benson as inventors. It was filed on 18 September 1998 and granted on 24 July 2001. The record provides an additional dated source for technical work on the approach (Linden, Jacobi and Benson, 2001).

The names also show why a history should not be reduced to the authorship of one later article. Different records can make different contributions visible. Publishing a paper does not establish that only its authors participated in every earlier stage of development.

A patent does not, however, determine the exact date of first operational deployment or the commercial effect of every described feature. It documents a procedure and a patenting event. Claims about implementation, public explanation and business outcomes still require evidence appropriate to each question.

This is historical source criticism rather than an assessment of present-day patent rights. It permits a specific milestone to be retained without turning it into a universal claim of invention. Such distinctions can be lost when a complex development is compressed into one line on a timeline.

Recommendation became a marketing task inside the store

J. Ben Schafer, Joseph Konstan and John Riedl’s 1999 study examined recommender systems across several e-commerce sites. It related recommendations to finding products, additional selling opportunities and customer relationships, while distinguishing interfaces, technologies and inputs. Recommendations were already being analysed as business tools rather than only unusual computing experiments (Schafer, Konstan and Riedl, 1999).

The marketing task was partly relocated within the offer. Not every commercial influence had to originate in an external advertisement. Choosing which products appeared in the shop could itself shape which possibilities a visitor considered. Selection and ordering became consequential decisions.

This did not abolish assortment management. A retailer still had to determine availability and provide useful product descriptions. Recommendation added another layer by selecting a smaller visible set from the larger stock of possibilities. That set did not have to be identical for everyone.

The relationship between product and communication also changed. An effective selection aid could simultaneously be a service and a route to further sales. Customer usefulness and the retailer’s commercial interest could align, but alignment was not guaranteed merely because the procedure was personalized. It depended on what the system actually helped the person do.

The 2003 account described selection, not generative AI

Linden, Smith and York’s 2003 paper discussed recommendations on web pages, in shopping baskets and in email campaigns. It reported higher click-through and conversion than untargeted content, but did not provide a publicly reproducible, isolated return-on-investment result for 1998 (Linden, Smith and York, 2003).

The system should therefore not be equated with present-day text generators. Selecting and ranking existing products does not require generating new product descriptions. “AI in marketing” refers here to a field of data-based inference and selection. The phrase should not imply capabilities that the historical application did not demonstrate.

The data description needs similar care. The possibility of using different signals does not prove that every early implementation processed every search, click and purchase in the same way. A historical account should identify documented inputs rather than making a blanket claim of complete behavioural knowledge.

The narrower explanation remains consequential. The innovation did not require knowing everything about every person. It involved using particular observed relationships to create a useful choice. Its effectiveness depended on the quality of that relationship, rather than on the breadth of the language used to describe artificial intelligence.

Observed behaviour was evidence, not a complete preference profile

A purchase might be intended for the buyer or selected as a gift. Viewing an item can indicate enthusiasm, comparison or curiosity. A missing rating need not express dislike. These examples expose an interpretive problem: the recorded event is not identical to the interest behind it.

More data does not necessarily eliminate every such ambiguity. People change interests and may purchase on behalf of several household members. A large dataset remains a collection of observable actions. The system must decide which actions to emphasize and how strongly.

Personalization can obscure this limitation. An interface feels individual because it responds to a recognizable history. That does not mean the system comprehensively understands the person. A limited predictive procedure can be useful without being a full account of human taste.

The experience also depends on how mistakes are encountered. An unsuitable suggestion may be accepted as an occasional error or interpreted as evidence of general unreliability. Technical output and perceived quality are separate, connected dimensions. Marketing outcomes depend partly on that human interpretation.

New items created a problem that calculation alone could not remove

Andrew Schein, Alexandrin Popescul, Lyle Ungar and David Pennock’s 2002 paper examined cold-start recommendations for items nobody in the community had yet rated. It combined content information with collaborative evidence. The study addressed a limit that a purely historical-rating approach could not solve simply by running faster (Schein et al., 2002).

A new book can fit somebody’s interests while lacking a recorded pattern connecting it with other titles. Commercial availability and recommendation evidence can therefore develop at different times. The gap is also a merchandising problem: entering the catalogue does not automatically make an item easy to recommend through every method.

New visitors present a related but distinct challenge. Without a recognizable history, there is less evidence for an individual selection. General starting suggestions, explicitly stated interests or current context are possible information sources. Whether a particular early implementation used them requires its own evidence; these possibilities do not establish additional Amazon features in 1998.

The dependency among assortment, audience and data becomes visible here. More products expand choice while initially lacking the relationships a procedure needs. More visitors may enrich evidence over time while still needing help on arrival. Personalization was not one switch that produced the same quality in every situation after it had been turned on.

Explanation was already a research question in 2000

Herlocker, Konstan and Riedl’s 2000 research investigated how and why collaborative-filtering recommendations should be explained. Their experiments provided evidence that explanations could improve acceptance. The study concerned the presentation of recommendations, not a measured Amazon revenue effect (Herlocker, Konstan and Riedl, 2000).

The issue matters commercially. An unfamiliar item may look inappropriate until a connection with an earlier selection becomes understandable. Explanation need not alter the computed order to change how the result is assessed. It supplies additional information for interpretation.

Greater acceptance is not automatically a better decision, however. A persuasive explanation can also make a weak suggestion attractive. Evaluating the interface therefore requires more than asking whether it encourages clicking. It also matters whether people can understand the basis of the recommendation. This is an analytical implication, not an additional reported result of the study.

The early research shows why computation and judgement cannot be separated completely. A calculated output still has to be translated into a usable interaction. The history of recommender systems includes that translation as well as the development of algorithms.

General popularity was different from distinctive relevance

Larry Hardesty’s 2019 Amazon Science retrospective discussed why frequent joint purchasing does not automatically demonstrate a particularly strong relationship. Very popular items appear in many histories. It also described later corrections concerning unusually active purchasers. The account reveals continuing work on the meaning of statistical association (Hardesty, 2019).

A hypothetical bestseller illustrates the point. If it appears in many orders, it will also occur alongside numerous specialist titles. That alone does not make it distinctive for every specialist audience. General frequency and an unusually strong connection are different statements.

The choice of measure affects what becomes visible. A procedure can follow its rule correctly while repeatedly foregrounding familiar products. Its suitability must therefore be judged against the intended task, rather than inferred from the mere fact that it operates mathematically.

For marketing history, this is more than a technical detail. Recommendation distributes attention within an assortment. How it handles popularity can help determine whether it supports discovery or reinforces existing visibility. The computational rule participates in the organization of demand.

Better prediction was not automatically greater business value

Jonathan Herlocker and his co-authors’ 2004 review distinguished tasks, datasets, metrics and user evaluations. It showed that recommender systems had been assessed in many, often difficult-to-compare ways. A single accuracy score could not express every relevant property (Herlocker et al., 2004).

A retrospective test might ask whether known preferences can be predicted. An operating retailer must also ask whether the displayed suggestions help at that moment. An accurately predicted item may already be familiar. A less predictable suggestion might provide useful discovery. The value of the property depends on the task.

Dietmar Jannach and Michael Jugovac’s 2019 review addressed business value directly. It identified continuing questions about quantifying real-world effects and about translating algorithmic improvements into commercial outcomes. The distinction between an improved offline test and demonstrated business value remains essential (Jannach and Jugovac, 2019).

For the 1998 milestone, deployment can be documented, as can later technical significance. A specific assertion about additional profit still needs a separate demonstration. Neither the company’s prominence nor an attractive algorithm substitutes for that evidence.

Click-through, conversion and return on investment asked different questions

A click-through rate relates interactions to impressions. A conversion rate relates defined outcomes to a specified starting population. Financial return also depends on costs, margins and outcomes that actually occurred because of the intervention. Without those distinctions, a precise percentage can support an imprecise historical narrative.

Consider an explicitly invented comparison. Under comparable conditions, one variant produces four purchases per hundred visits and another five. That is initially one additional purchase in the group, not proof of an equivalent gain in profit. Basket value, returns, displacement of other buying and operating costs still matter. These are illustrative numbers, not Amazon results.

A purchase after a recommendation click is not necessarily caused wholly by the recommendation. A customer may already have intended to acquire the product or might have found it through another route. Estimating the additional effect requires a suitable comparison. The observed sequence of events is insufficient by itself.

This is why a blanket statement that recommendations proved higher marketing ROI in 1998 goes beyond the evidence unless an appropriate study is identified. A careful claim remains useful: a powerful form of personalized selection developed, and its different effects require different kinds of measurement.

The historical issue is not whether the technology could be commercially useful. It is whether a particular outcome, magnitude and date have actually been established. Maintaining that distinction avoids turning a credible innovation into an unsupported success statistic.

Personal discovery did not guarantee greater market-wide diversity

Daniel Fleder and Kartik Hosanagar’s 2009 study used analytical modelling and simulation to investigate sales diversity. It showed how individual discovery and the distribution of aggregate demand could diverge. Under particular conditions, recommendations could broaden personal consumption while concentrating overall sales (Fleder and Hosanagar, 2009).

The distinction is intuitive. Many people can each discover something new to them while discovering the same already popular product. Their personal experience has expanded, but demand across the market has not necessarily become more evenly distributed. Individual novelty and collective diversity operate at different levels.

The finding does not prove that Amazon’s 1998 system displaced every niche. It also prevents the opposite automatic conclusion that recommendations must always promote the entire breadth of a catalogue. Actual effects depend on the procedure, audience and initial distribution.

For international marketing history, this leaves an empirical question. A method may operate differently across assortments and patterns of use. A prominent American application does not supply a complete description of every subsequent deployment around the world.

What remains of Amazon’s milestone

The year 1998 is a useful reference point for Amazon’s introduction of item-based collaborative filtering, not for the beginning of all personalization. The background includes Tapestry, GroupLens, Ringo and hybrid approaches. The company report records earlier Amazon features, the patent documents 1998 technical work and the 2003 article explains a later publicly described implementation.

For marketing, the significant change was that selection within a shop could become an ongoing computational service. A large catalogue did not have to be presented identically to everyone. Recorded evidence could help identify possibilities that a visitor might otherwise overlook, making presentation part of the offer itself.

The significance does not require an unsupported revenue percentage. Data could influence what appeared next, rather than only describing past activity. A computational procedure stood between human interests and a commercial interface. Its relevance, usefulness and financial contribution remained questions requiring concrete evidence.

References

Amazon.com (1998) Form 10-K for the year ended December 31, 1997, especially ‘Recommendations and Personalization’. https://s2.q4cdn.com/299287126/files/doc_financials/annual/123197_10k.pdf.

Balabanović, M. and Shoham, Y. (1997) ‘Fab: content-based, collaborative recommendation’, Communications of the ACM, 40(3), pp. 66–72. https://doi.org/10.1145/245108.245124.

Fleder, D. and Hosanagar, K. (2009) ‘Blockbuster Culture’s Next Rise or Fall: The Impact of Recommender Systems on Sales Diversity’, Management Science, 55(5), pp. 697–712. https://doi.org/10.1287/mnsc.1080.0974.

Goldberg, D., Nichols, D., Oki, B.M. and Terry, D. (1992) ‘Using collaborative filtering to weave an information tapestry’, Communications of the ACM, 35(12), pp. 61–70. https://doi.org/10.1145/138859.138867.

Hardesty, L. (2019) ‘The history of Amazon’s recommendation algorithm’, Amazon Science, 22 November. https://www.amazon.science/the-history-of-amazons-recommendation-algorithm.

Herlocker, J.L., Konstan, J.A. and Riedl, J. (2000) ‘Explaining collaborative filtering recommendations’, Proceedings of CSCW ’00, pp. 241–250. https://doi.org/10.1145/358916.358995.

Herlocker, J.L., Konstan, J.A., Terveen, L.G. and Riedl, J.T. (2004) ‘Evaluating collaborative filtering recommender systems’, ACM Transactions on Information Systems, 22(1), pp. 5–53. https://doi.org/10.1145/963770.963772.

Jannach, D. and Jugovac, M. (2019) ‘Measuring the Business Value of Recommender Systems’, ACM Transactions on Management Information Systems, 10(4), Article 16, pp. 1–23. https://doi.org/10.1145/3370082. Author version: https://arxiv.org/abs/1908.08328.

Linden, G.D., Jacobi, J.A. and Benson, E.A. (2001) Collaborative recommendations using item-to-item similarity mappings, US Patent 6,266,649, filed 18 September 1998, granted 24 July 2001. https://patents.google.com/patent/US6266649B1/en.

Linden, G., Smith, B. and York, J. (2003) ‘Amazon.com recommendations: item-to-item collaborative filtering’, IEEE Internet Computing, 7(1), pp. 76–80. https://doi.org/10.1109/MIC.2003.1167344. Full text: https://www.cs.umd.edu/~samir/498/Amazon-Recommendations.pdf.

Resnick, P., Iacovou, N., Suchak, M., Bergstrom, P. and Riedl, J. (1994) ‘GroupLens: an open architecture for collaborative filtering of netnews’, Proceedings of CSCW ’94, pp. 175–186. https://doi.org/10.1145/192844.192905.

Resnick, P. and Varian, H.R. (1997) ‘Recommender systems’, Communications of the ACM, 40(3), pp. 56–58. https://doi.org/10.1145/245108.245121.

Sarwar, B., Karypis, G., Konstan, J. and Riedl, J. (2001) ‘Item-based collaborative filtering recommendation algorithms’, Proceedings of WWW ’01, pp. 285–295. https://doi.org/10.1145/371920.372071.

Schafer, J.B., Konstan, J. and Riedl, J. (1999) ‘Recommender systems in e-commerce’, Proceedings of EC ’99, pp. 158–166. https://doi.org/10.1145/336992.337035.

Schein, A.I., Popescul, A., Ungar, L.H. and Pennock, D.M. (2002) ‘Methods and metrics for cold-start recommendations’, Proceedings of SIGIR ’02, pp. 253–260. https://doi.org/10.1145/564376.564421.

Shardanand, U. and Maes, P. (1995) ‘Social information filtering: algorithms for automating “word of mouth”’, Proceedings of CHI ’95, pp. 210–217. https://doi.org/10.1145/223904.223931.

Smith, B. and Linden, G. (2017) ‘Two Decades of Recommender Systems at Amazon.com’, IEEE Internet Computing, 21(3), pp. 12–18. https://doi.org/10.1109/MIC.2017.72. Full text: https://cdn.amazon.science/76/9e/7eac89c14a838746e91dde0a5e9f/two-decades-of-recommender-systems-at-amazon.pdf.

Amazon Associates became a defining example in the history of affiliate marketing. Its significance does not depend on proving that Amazon invented commissions or every earlier online referral arrangement. The important question is how a retailer connected many independent websites with a common transaction and fulfilment system. Selection and audience relationships could remain decentralized while orders and delivery were handled centrally. That combination made a recommendation part of a larger commercial network.

What changed in the division of work

Affiliate marketing is used here to mean an arrangement in which a partner refers visitors to a provider and receives an agreed payment when a qualifying action occurs. In the early Amazon case, the relevant action was a referred sale. A link provided the technical connection, while programme rules determined which transaction could be attributed and rewarded. The economic institution was therefore more than a clickable reference.

An editorial link can help readers find useful information. An affiliate link may do that while also participating in a payment relationship. The additional relationship changes the recommender’s incentives. Publication becomes connected with distribution: the website can benefit economically when its guidance leads to a purchase.

That did not necessarily make the partner a retailer in the same sense as Amazon. Recommending a book did not require performing every task involved in supplying it. Nor did programme participation automatically make the website owner an employee. The relationship could connect independent organizations without turning every participating site into a branch shop.

Later professional analysis reflected this relational character. Dennis L. Duffy’s 2005 article described the affiliate marketplace and emphasized mutual benefit between advertiser and affiliate. It provides a conceptual interpretation of the model, rather than evidence for every detail of Amazon’s launch nine years earlier (Duffy, 2005).

Why Amazon cannot simply be named the sole inventor

In November 2000, Shawn Collins challenged the idea that Amazon’s programme had no predecessors. His near-contemporary industry history named PC Flowers & Gifts and early activity by CDnow, among other examples. The account establishes that the origin story was already disputed. Its dates and participant statements should not be treated as a retrospectively complete chronology of every referral arrangement (Collins, 2000).

The difficulty begins with the category. An individual commission agreement, a programme involving several partners and a broadly accessible technical system are different developments. Earlier online services and the open web also provided different environments. Grouping all these arrangements under a present-day label can conceal the distinctions that explain their operation.

A narrower claim is stronger: Amazon Associates was an early and widely adopted web-retailing partner programme whose expansion can be followed through contemporary company records. That proposition does not require an inventor’s crown. It can be investigated through the work assigned to each party, reported enrolment and changes in incentives.

“Scalable” should not become a substitute for evidence of priority. The word describes the ability to extend relationships through a reusable procedure. Establishing that nobody else possessed a similar capability earlier would require a separate inquiry. Amazon’s rapid expansion is documented; absolute precedence is a different question.

What the July 1996 announcement actually establishes

Amazon reported more than 300 websites. Partners selected books and linked to Amazon, which handled ordering, payments, service and shipping. This documents the proposed division of work, not every participant’s profitability (Amazon.com, 1996).

The variety of interests explains the commercial logic. A general bookseller cannot necessarily reproduce every specialist’s relationship with an audience. An independent publisher may understand the questions its readers ask and the level of material they need. The arrangement could connect that knowledge with a much broader retail offering.

This does not establish that every recommendation was useful or independent. It identifies the proposition Amazon was promoting. A publisher already serving an audience could add a route to a relevant purchase, while the retailer could reach people outside its own homepage. The two organizations contributed different resources.

The announcement should also be distinguished from the creation of every individual relationship. A programme may already have participants when it is publicly presented. Reported enrolment does not identify the exact moment of the first attributed purchase. It does provide a sound basis for treating July 1996 as a significant point in the programme’s history.

A broad catalogue could be interpreted through many specialist contexts

Books offer numerous opportunities to connect information with purchasing. An article can point readers towards further study. A specialist can explain whether a title suits beginners or experienced practitioners. The value of the recommendation lies in selection and interpretation, not simply in the existence of a link.

The arrangement joined two different systems of organization. A retailer maintained a catalogue of items available to purchase. Partners could interpret that catalogue through the interests of particular audiences. One title could answer different questions in different settings. Decentralized explanation complemented centralized availability.

This also created a possible role for smaller publications. A site with limited overall traffic might still be closely aligned with a particular interest. Audience size alone did not determine the relevance of a recommendation. A network containing many specialist contexts could consequently be useful even if no individual small site generated many orders.

That is an analysis of the model, not a recovered sales distribution. Membership totals do not show how transactions were divided among large and small participants. They establish an expanding enrolment base. The economic value created at each point would need separate evidence.

In 1997, an accessible programme acquired selective incentives

On 8 September 1997, Amazon reported more than 15,000 associate sites and offered a temporary incentive for the top 500 PC Meter sites. The announcement specified enhanced referral rates of 22.5 per cent for one group of titles and 7.5 per cent for another, on qualifying sales between October 1997 and March 1998 (Amazon.com, 1997).

The detail challenges the image of a completely uniform programme. Broad accessibility and special arrangements for major participants could coexist. Amazon sought not only additional partners but particular kinds of reach. Expansion involved selective recruitment as well as the repeated enrolment of small websites.

The historical percentages belong to that promotion. They should not be presented as universal rates for every book, or as present-day conditions. A maximum payment also says little about an average participant’s earnings. The included products and the definition of a qualifying transaction matter.

For marketing history, the episode illustrates how standardization and differentiation could operate together. A recognizable technical relationship could support distinct economic offers. The programme therefore required decisions about whom to recruit and which incentives to use, as well as a method for accepting applications.

The 1998 membership figures show expansion, not income

In February 1998, Amazon reported more than 30,000 member websites and described special relationships with major services including AOL.com, Yahoo!, Netscape, GeoCities and Excite. The account documented both growing enrolment and strategic partnerships. These were related dimensions of expansion, but not necessarily identical commercial arrangements (Amazon.com, 1998a).

In June, Amazon reported more than 60,000 participants. It described free enrolment, fees of up to 15 per cent and weekly sales reports by email. Adobe’s connection between software information and relevant books illustrated the thematic fit being promoted (Amazon.com, 1998b).

An enrolment total is not an income statement. It leaves open how many members remained active, how many referred sales and how revenues were concentrated. It does not reveal whether an individual publisher covered its production costs. A registered website and a profitable publishing business are different statistical units.

The reporting arrangement nevertheless deserves attention. A partner needed some way of understanding whether its references produced credited results. Reporting was part of the service, not merely administration after the event. A repeatable programme required feedback as well as enrolment.

The significance of the figures is therefore specific. They show that Amazon publicly described a rapidly widening partner base. They cannot, by themselves, establish the total incremental revenue generated by the programme or the distribution of benefits within it.

Music widened the range of relevant publications

In August 1998, Amazon announced more than 100,000 members and highlighted music participants including Spin, Vibe and Spinner.com. The release connected Associates with the expansion from books into music. Different publishing environments were becoming part of the programme’s commercial story (Amazon.com, 1998c).

An additional category does more than increase the item count. It can make participation relevant to publishers whose existing material fits less closely with the original assortment. A music publication can create different referral occasions from a cooking site. A common procedure can operate within distinct cultural contexts.

The content requirements still differ. Explaining the suitability of a technical book is not the same task as discussing a recording. A link may have a similar technical structure while the recommendation involves a different kind of judgement. Standardizing the transaction did not standardize every form of communication.

This was an important feature of the division of labour. The retailer could concentrate transaction capabilities without writing every recommendation. Independent participants contributed styles, expertise and relationships that varied. Their usefulness remained a separate question from the technical functioning of the programme.

The annual report reveals the less visible half of the system

Amazon’s report for 1997 described Associates as extending its market presence. The same document discussed procurement, inventory and order fulfilment, including purchasing from publishers and distributors and the importance of Ingram. The visible referral link was connected with a much larger retail organization (Amazon.com, 1998d).

This guards against an explanation based only on communication. A recommendation can create interest, but the promised product must subsequently be supplied. Availability, payment handling and delivery help determine whether the customer has a satisfactory experience. Growth in referrals depends partly on capabilities behind the visible purchase process.

The case also complicates a simple disintermediation narrative. Independent recommenders were being added to the commercial system while established supply relationships remained important. Some tasks were concentrated; others were connected to external organizations. The result was a different combination of intermediaries rather than the elimination of every intermediate function.

The website changed access to a transaction without making physical procurement disappear. Affiliate marketing was a route to customer acquisition, but its value depended on an actual retail service. The distinction explains why a persuasive recommendation and a functioning link could not, by themselves, complete the business.

International reach was not the same as local commercial readiness

In October 1998, Amazon announced German and British websites replacing operations acquired earlier that year from ABC Bücherdienst and Bookpages. The release emphasized local availability and improved shipping arrangements for many US titles. It illustrates the difference between international technical accessibility and a retail offer adapted to particular markets (Amazon.com, 1998e).

That distinction matters to the referral model. A publisher may attract visitors from several countries, but those contacts do not necessarily present equivalent opportunities. Language, assortment, shipping costs and delivery times can affect the practical value of a recommendation. Crossing a border with a link does not solve every condition of purchasing.

The European expansion should therefore not be projected backwards into a fully developed global programme in July 1996. Nor should opening a local store automatically be treated as the exact launch date of its national affiliate arrangement. Those are different events requiring different evidence.

The history consequently involves more than exporting an unchanged American idea. Local businesses, audiences and supply capabilities helped determine whether the offer made sense in another environment. An international medium did not erase the organizational work needed to serve an international customer base.

Patents document particular technical claims and dates

William J. Tobin’s US Patent 6,141,666 records a priority date of 22 January 1996, an application on 21 January 1997 and a grant on 31 October 2000. Priority and the later application are distinct dates. The record documents a particular technical proceeding, not every broad claim made about the first use of affiliate marketing (Tobin, 2000).

Amazon’s US Patent 6,029,141, titled “Internet-based customer referral system,” was filed on 27 June 1997 and granted on 22 February 2000. Jeff Bezos is among the named inventors. Those dates should not be conflated with the programme’s July 1996 launch (Bezos et al., 2000).

Patents are valuable historical sources when the question is specified carefully. They identify people, dates and described procedures. They do not automatically prove that nobody operated a related business arrangement earlier. Nor does a grant, without further analysis, show that every later partner programme used the same technical claim.

The purpose here is historical source criticism rather than an assessment of present-day rights. Technical documentation and narratives of commercial pioneering can overlap without answering identical questions. Separating them permits the documents to be used without making them settle an entire origin story.

In 2000, the programme placed an explicit value on new customers

In August 2000, Amazon reported 500,000 enrolled associates and announced temporary new-customer bonuses. The thresholds began at ten new customers referred per quarter, with larger payments for higher totals. The promotion highlighted a distinction that can disappear when only referred order value is considered: different transactions can serve different acquisition goals (Amazon.com, 2000).

A retailer may value an existing customer’s additional order differently from a first order by someone not previously reached. The distinction concerns the expected relationship beyond the immediate basket. A bonus can direct partners towards the type of result the retailer particularly wants.

The episode also shows that a programme required continuing design. Its technical structure could persist while economic incentives changed. Growth objectives were translated into conditions for additional payments. Affiliate marketing developed through those commercial decisions as well as through the improvement of linking technology.

Again, half a million registrations should not be interpreted as half a million equally successful businesses. The number describes enrolment reported by the company. It does not establish that every participant made meaningful income. This matters when a large historical network is later presented as evidence of an easy earning opportunity.

The published schedule makes the incentives more concrete. It offered US$20 for ten to nineteen new customers. At one thousand, the payment was US$5,000, plus five dollars for each additional new customer. Those amounts belonged to the temporary 2000 promotion and its other conditions (Amazon.com, 2000).

A schedule with thresholds differs from an identical payment for every purchase. Near a threshold, the economic attraction of another qualifying introduction can change. A participant approaching the next level might value an additional referral differently from before. That does not establish that every partner changed its behaviour in this way. It identifies a possibility created by the rule.

The distinction is historically useful because it reveals the programme as a designed incentive system. The shared infrastructure did not merely record movement between websites. It communicated which outcomes the retailer particularly wanted and attached a financial consequence to achieving them. The bonus schedule provides more precise evidence of that priority than a generic claim that all digital marketing sought more traffic.

The total value of a customer’s eventual relationship remained uncertain. A published acquisition bonus does not reveal how profitable every acquired customer became. It shows what the company was prepared to offer under a particular campaign. Price, expected future value and realized outcome are different quantities, even when they are connected within one marketing decision.

Recommender, supplier and seller were different roles

The proximity between content and purchasing could make every participating website look like a complete shop. Analytically, the roles should remain separate. A recommender connects an audience with an offer. A supplier provides goods. A seller performs the relevant role within the purchase relationship. One organization may undertake several tasks without making the tasks interchangeable.

A hypothetical author illustrates the distinction. On her website, she might recommend another person’s book and receive a fee for a qualifying referral. She might also supply or sell her own books under a different arrangement. The site’s appearance alone would not establish the commercial relationship behind it. The example explains possible roles; it is not an additional documented Amazon contract from 1996.

The limited role was part of the model’s attraction. A publisher could participate in a transaction without building every retail capability. That limitation also created dependence on another organization. A persuasive recommendation could not ensure that the later order would proceed without difficulty.

The distinction explains which work remained with the publisher, which belonged elsewhere and why the parties depended on one another. The model redistributed functions. It did not make those functions disappear.

Performance-based payment redistributed risk

A fee linked to an attributed sale differs from buying an opportunity for exposure. The merchant pays the relevant commission after the defined event. The partner has a stronger interest in turning attention into a qualifying action. Creating and maintaining the content, however, can require work before any sale occurs.

An explicitly hypothetical calculation clarifies the relationship. Referring 1,000 units of qualifying sales value at a five per cent fee would produce 50 units before the publisher’s own costs. Whether that rewards the work adequately depends on production, maintenance and other expenses. Five per cent is an illustrative rate, not a statement of historical or current Amazon conditions.

For the retailer, commission is not the only expense either. Operating the business and fulfilling orders still require resources. Performance pricing can relocate one part of acquisition risk without removing other costs. Describing the arrangement as universally risk-free obscures that distinction.

Attribution should also remain separate from causation. A purchase can be credited to a partner under an agreed rule without proving that it would never have occurred otherwise. The allocation rule resolves an operational question about payment. The amount of additional business caused by the activity is another question. This is an analytical boundary of the model, not a claim about an undocumented Amazon experiment.

Trust became part of the commercial relationship

A publisher can build credibility over time. A commercial recommendation may draw on that relationship. A relevant suggestion can help readers, while a selection driven only by commission can overlook their needs. The interests of the audience and the economic incentive of the recommender do not automatically coincide.

The original division of work explains the tension. A merchant does not write every statement on an independent partner’s site. The partner does not control every aspect of the subsequent customer experience. Both can benefit from a successful introduction, while each may be affected by the other’s failures.

This is not an allegation that early participants systematically behaved badly. It follows from cooperation among separate organizations. Technology must be accompanied by expectations about acceptable practice and reliable accounting. As the network grows, personal familiarity with every participant becomes less feasible.

The relationship therefore cannot be understood solely through a percentage. Selection quality, audience confidence and dependable fulfilment contribute to whether the arrangement remains useful. A system that successfully records transactions still needs a reason for people to trust the recommendations that initiate them.

A merchant’s programme and an affiliate network occupied different positions

An Internet Capital Group release filed with the SEC in 2005 documented the sale of LinkShare to Rakuten. It gave 1996 as LinkShare’s founding year and described technology, reporting and management services for performance-based partnerships. The record shows another organizational layer: arranging the referral relationship could become a separate business (Internet Capital Group, 2005).

A merchant can operate a programme for its own assortment. A service provider can supply facilities through which several merchants and publishers cooperate. In the first case, the central purpose is acquisition for the merchant. In the second, organizing such relationships for other businesses becomes the offer.

An additional intermediary can consequently emerge from a technology associated with direct access. Accounting, reporting and supervision need not be developed independently by every merchant. Choosing which capabilities to operate internally and which to obtain from specialists becomes part of the market’s structure.

The Japanese acquisition of an American provider also broadens the international perspective. The later history did not consist only of one US retailer extending its own programme abroad. The infrastructure suppliers themselves became involved in cross-border corporate relationships. This is an outlook on institutional development, not a reason to treat LinkShare as part of Amazon Associates or to assign all programmes one origin.

Later research made supervision and incentives visible

Benjamin Edelman and Wesley Brandi’s 2015 research distinguished clear affiliate violations from problematic borderline practices. External specialists were more effective at excluding the former, while internal staff did better with the latter. The study concerns a later market and does not establish specific conduct by Amazon’s 1996 partners (Edelman and Brandi, 2015).

Its historical relevance is nevertheless clear. Operating a large network requires more than technical attribution. Someone must judge which practices actually serve the merchant’s interests and decide who will enforce the rules. Specialist expertise and aligned incentives can provide different advantages.

Scalability therefore has more than one dimension. Enrolling many participants is one task; maintaining the desired quality across them is another. A durable programme needs relationships that are understandable and governable as well as numerous. The institutional work behind the link belongs within marketing history.

The later findings should not be used to portray every early programme as already possessing a mature solution. They identify problems that help explain why management practices developed around the original commercial proposition. The wider the network, the more significant those organizational choices could become.

What Amazon Associates changed in marketing history

The 1996 programme connected a central retailer with decentralized recommendation settings. It could use the diversity of existing websites rather than building every relevant context itself. Its documented development from specialist participants to major web partnerships and hundreds of thousands of registrations illustrates an influential route to expansion.

The achievement should remain separate from other innovations. An externally rewarded recommendation is not the same as an algorithm selecting products inside a shop. Payment for a qualifying referral differs from purchasing a banner placement or a search advertisement. These arrangements can coexist while retaining their own rules and development histories.

The defensible conclusion is not that Amazon invented every form of affiliate marketing. It made a related model exceptionally visible and widely adopted within its retail business. Its significance lies in connecting specialist selection, repeatable participation, attributed payment and centralized fulfilment.

The link became an interface between independent organizations and an audience. A recommendation could contribute to distribution without requiring every recommender to operate a complete retail business. That reorganization of work explains Amazon Associates’ historical importance more precisely than a simple claim to have been first.

References

Amazon.com (1996) ‘Web Sites Big and Small Participating — From Film Critics to Chefs to Puppies’, release text dated 18 July; archive header dated 17 July. https://press.aboutamazon.com/1996/7/web-sites-big-and-small-participating-from-film-critics-to-chefs-to-puppies.

Amazon.com (1997) ‘Earth’s Biggest Bookstore Offers Premium Associate Incentive to Top 500 PC Meter Sites’, 8 September. https://press.aboutamazon.com/1997/9/earths-biggest-bookstore-offers-premium-associate-incentive-to-top-500-pc-meter-sites.

Amazon.com (1998a) ‘Amazon.com Associates Program Surpasses 30,000 Members’, release text dated 18 February; archive header dated 17 February. https://press.aboutamazon.com/1998/2/amazon-com-associates-program-surpasses-30-000-members.

Amazon.com (1998b) ‘Amazon.com Associates Program Surpasses 60,000 Members’, release text dated 8 June; archive header dated 7 June. https://press.aboutamazon.com/1998/6/amazon-com-associates-program-surpasses-60-000-members.

Amazon.com (1998c) ‘Amazon.com’s Associates Program is a Hit with Music Sites’, release text dated 3 August; archive header dated 2 August. https://press.aboutamazon.com/1998/8/amazon-coms-associates-program-is-a-hit-with-music-sites.

Amazon.com (1998d) Form 10-K for the year ended December 31, 1997, especially Item 1: Business. https://s2.q4cdn.com/299287126/files/doc_financials/annual/123197_10k.pdf.

Amazon.com (1998e) ‘Amazon.com Enters European Book Market’, release text dated 15 October; archive header dated 14 October. https://press.aboutamazon.com/1998/10/amazon-com-enters-european-book-market.

Amazon.com (2000) ‘Amazon.com Associates Program Celebrates 500,000 Member Milestone With New-Customer Referral Bonuses’, release text dated 3 August; archive header dated 2 August. https://press.aboutamazon.com/2000/8/amazon-com-associates-program-celebrates-500-000-member-milestone-with-new-customer-referral-bonuses.

Bezos, J.P., Kaphan, S.J., Ratajak, E.L. and Schonhoff, T. (2000) Internet-based customer referral system, US Patent 6,029,141, filed 27 June 1997, granted 22 February 2000. https://patents.google.com/patent/US6029141A/en.

Collins, S. (2000) ‘History of Affiliate Marketing’, ClickZ, 10 November. https://clickz.com/history-of-affiliate-marketing/76414/.

Duffy, D.L. (2005) ‘Affiliate marketing and its impact on e-commerce’, Journal of Consumer Marketing, 22(3), pp. 161–163. https://doi.org/10.1108/07363760510595986.

Edelman, B. and Brandi, W. (2015) ‘Risk, Information, and Incentives in Online Affiliate Marketing’, Journal of Marketing Research, 52(1), pp. 1–12. https://doi.org/10.1509/jmr.13.0472.

Internet Capital Group (2005) ‘ICG Announces Completion of Sale of LinkShare Corporation to Rakuten, Inc.’, 3 October, SEC Exhibit 99.1. https://www.sec.gov/Archives/edgar/data/1085621/000119312505195907/dex991.htm.

Tobin, W.J. (2000) Method and system for customizing marketing services on networks communicating with hypertext tagging conventions, US Patent 6,141,666, priority 22 January 1996, filed 21 January 1997, granted 31 October 2000. https://patents.google.com/patent/US6141666A/en.

The historical achievement was consequently more substantial than an uncontested first place. A publisher, a telecommunications brand and creative practitioners explored how a graphic advertisement could also become a direct route to another experience. The image was both a message and an entrance. A familiar method of media finance acquired an unfamiliar interface, raising questions about what could be counted and what those counts could establish.

Why the phrase “first banner ad” needs qualification

The HotWired launch provides a firm reference point, but not a simple birthday for all online advertising. Digiday’s later oral history describes a launch involving advertisements from several brands. The familiar account of one AT&T banner therefore condenses a collective event. The interviews are valuable participant testimony, rather than a complete technical record capable of settling every claim about which file appeared first (Digiday, 2017).

“First” can mean the first commercial communication on the internet, the first advertisement on a website, the first linked graphic or an early, widely recognized sale of a repeatable banner format. These categories are not interchangeable. A timeline can give a plausible date while attaching an excessively broad claim to it. Historical accuracy requires specifying the activity being dated.

A more defensible description is that HotWired helped make graphical web advertising visible and purchasable for major brands, with AT&T supplying its most memorable early example. This preserves the importance of the event without claiming that commercial communication had never appeared online before. Predecessors do not diminish the achievement. They clarify which particular combination became influential.

Online advertising is also a broader category than web banners. The internet includes services other than websites, while a commercial message does not have to be a clickable image. The HotWired case is clearest when the focus remains on graphical advertising within the open web. Other histories of electronic promotion retain their own chronology.

The distinction matters because origin stories shape explanations of later success. If the entire industry is imagined as the unfolding of one image, the work of publishers, agencies, researchers and standards organizations disappears. If the banner is placed within a developing commercial environment, the later institutions become visible as achievements that also required decisions and experimentation.

The infrastructure and the commercial precedents came earlier

The technical foundations did not originate in an advertising agency. On 30 April 1993, CERN placed key components of its web software in the public domain. Its account identifies the basic client, server and shared code library. This release supported the wider dissemination of an infrastructure on which many kinds of public and commercial activity could be developed (CERN, n.d.).

The decision did not invent an advertising format. It helped establish conditions in which organizations could build services without the web being a single publisher’s proprietary publication. This gives the story an international dimension. An important commercial experiment in the United States relied on an infrastructure whose development and release were closely associated with a European research institution.

There were commercial web precedents as well. O’Reilly identifies the Global Network Navigator, established in 1993, as an early advertising-supported web portal. The Computer History Museum likewise describes GNN as a commercial site combining embedded advertising with different kinds of information. These accounts establish a history before HotWired without settling every narrower question about the first banner image (O’Reilly Media, n.d.; Computer History Museum, n.d.).

Making a link technically possible and persuading a large brand to buy a placement were different tasks. Commercial participation needed an intelligible offer, a design and some account of the value being purchased. The market did not yet possess the long history of comparable transactions that later buyers could consult. HotWired’s importance lies partly in its response to that organizational problem.

This is a recurring distinction in business history. A capability can exist before a convincing way of selling it becomes established. The creation of an influential commercial format requires more than demonstrating the capability. It also requires translating it into the language of customers, budgets and repeatable relationships.

Selling participation in a future with an uncertain audience

In the oral history, Jonathan Steuer recalled a price of US$10,000 a month for advertising within a HotWired section. Andrew Anker emphasized that no traffic commitments were made. The proposal included the prospect of publicity for brands willing to experiment. These are retrospective accounts of the sales approach, not a complete surviving set of contracts (Digiday, 2017).

The underlying logic is revealing. A publisher had to describe something unfamiliar in a form that an advertiser could authorize. Relating a web placement to established publishing prices offered a bridge. A buyer accustomed to magazine advertising could understand paying for a presence in an editorial environment, even if the audience and response remained uncertain.

The link nevertheless introduced a different possibility. The placement could lead somewhere else immediately. Advertisers were buying an opportunity to be noticed and a potential transition into a further experience. Those two possibilities were related but distinct. The first concerned exposure within the publication; the second concerned what happened when a person chose to leave it.

For a pioneering participant, the symbolic value of joining the experiment could matter too. Appearing in the new medium might support a reputation for technological imagination. Publicity about an advertisement could reach people who never encountered the original placement. Such effects should be distinguished from the advertisement’s direct influence on its viewers, because they require different evidence.

The early purchase should therefore not be described as though it were already a routine acquisition of a guaranteed number of clicks. A time-based presence and a performance-priced interaction assign uncertainty differently. Keeping that difference visible explains both the appeal and the difficulty of the initial commercial proposition.

“You Will” connected the banner with an existing campaign

McCambley explained in 2014 that AT&T’s banner drew on its established “You Will” campaign. The television advertising imagined future technological experiences, while the familiar words connected those expectations with a new action in a browser. The banner was thus related to the brand’s wider communication rather than an isolated digital curiosity (Digiday, 2014).

This complicates a common account of digital innovation as a complete break with previous media. Here, continuity helped make the unfamiliar intelligible. A recognizable campaign idea supplied a frame for an activity that many visitors had not yet learned to associate with advertising. The new medium offered a different way to enact an existing promise.

In television, the brand could show a possible future. In the browser, the visitor could take a small action immediately. The technology did not simply carry another representation of connection; it allowed the message to become an actual route. That was a creative translation between media, rather than merely the shrinking of a television idea into a narrow image.

The evidence should not be stretched into a claim about a fully integrated, precisely measured customer journey. What is documented is the connection between the campaign and the banner’s creative idea. It does not establish complete tracking of every contact across media. Recognizing that limit leaves the central achievement intact: a brand theme could remain familiar while adapting to a different mode of participation.

The case is useful because it shows that an apparently revolutionary execution can depend on accumulated meanings. A new communication channel does not necessarily make older investment irrelevant. The creative problem is to identify what can travel across media and what must change when the audience gains a different set of actions.

The destination offered culture rather than an order form

Digiday’s interview-based reconstruction in 2013 describes work by McCambley and Craig Kanarick on a destination connecting visitors with museum resources. It also reports US$30,000 for a three-month placement. These particulars come from participant recollections and should retain that attribution (Digiday, 2013).

The museum connection is more than a charming historical detail. A telecommunications brand’s proposition concerns connections whose value becomes understandable through what people can reach. Cultural resources supplied a reason to explore. Instead of making a technical specification the entire message, the experience could demonstrate a possible benefit of networked communication.

The sequence matters. The banner first had to attract curiosity, then make an action intelligible, then provide a destination that justified the invitation. Its effectiveness was therefore not wholly contained within the visible strip. The experience after the click formed part of the encounter. A compelling image followed by a disappointing destination would have created a different impression.

The arrangement can be interpreted as an early combination of promotion and useful guidance. That does not mean AT&T invented every practice later grouped under content marketing. Resemblance to a present-day category is not proof that the complete contemporary practice already existed. It does show that providing an experience of value was part of this early commercial experiment.

It also distinguishes an invitation from a completed sale. A person could benefit from the destination without purchasing telecommunications services. That possibility does not make the contact commercially meaningless. It means that an account of its value must consider what the brand intended to demonstrate and cannot simply count orders as though they were the only plausible outcome.

A small graphic performed several functions

The Web Design Museum’s record dates the AT&T object to 27 October 1994 and gives its dimensions as 476 by 56 pixels. This specific preserved object should not automatically be equated with dimensions subsequently treated as general banner standards. An artifact description and a later account of an industry format answer different questions (Web Design Museum, n.d.).

The creative idea makes the act of clicking its subject. Direct address introduces a small challenge; the arrow links the question with the promised answer. The short text leaves little reading to be done before acting. The advertisement helps explain its own use through the message it delivers.

That differs from a print advertisement whose response must occur somewhere else. The banner could begin the requested action within its own surface. Design had to make a commercial message noticeable while also making an unfamiliar interactive object understandable. The interface and the copy were not separate problems.

Simplicity, however, is not a universal performance formula. A prompt that is intriguing when a format is unfamiliar can become routine once audiences encounter similar prompts repeatedly. The historical explanation depends partly on the relationship between the design and the expectations of its users. It does not demonstrate that every short message with an arrow will work equally well.

The case also illustrates why preserving digital objects matters. A verbal description can convey the idea, but details of layout and dimensions help distinguish a particular execution from a generic later illustration. Historical interpretation benefits from knowing whether the subject is an actual early asset, a reproduction or a standardized format that became familiar afterwards.

A launch in 1994 was not the completion of standardization

On 10 December 1996, the Internet Advertising Bureau and CASIE announced voluntary recommendations covering eight commonly accepted banner sizes. The announcement linked the proposal to the proliferation of dimensions and the production burden on agencies. This is direct contemporary evidence of an organized standardization effort after the HotWired launch (Internet Advertising Bureau and CASIE, 1996).

Two achievements should therefore be kept apart. Publishing a recognizable example helps establish a format’s visibility. Coordinating shared specifications helps make it reusable across organizations. Standardization concerns relationships among sellers, buyers and producers, rather than only the appearance of one image. It provides common expectations for work that otherwise requires repeated negotiation.

For an agency, shared dimensions can reduce the amount of adaptation needed for different placements. For a publisher, they simplify the description of available space. For a buyer, they can make offers easier to compare. The economic importance lies partly in reducing coordination costs. A modest technical convention can help turn bespoke projects into a more repeatable market.

Calling the recommendations voluntary is important. They were not an inevitable law of screen design, nor did they require the disappearance of every other form. They were a response to a commercial problem identified by industry participants. That makes them an institution with a history, rather than a timeless property of the web.

HotWired can consequently be credited with helping establish banner advertising without being credited with completing the entire later standards process in 1994. The distinction makes the development more intelligible. A prominent example and a shared industry specification play complementary roles, but neither should be substituted for the other.

Measurement began with technical traces rather than dashboards

The 2013 reconstruction describes manual work with server logs. John Nardone recalled counting requests associated with the advertisement’s image file. The same account attributes the often repeated 44 per cent click-through figure to a participant. These details provide evidence of remembered practice, not an independently audited dataset with a fully disclosed calculation (Digiday, 2013).

A file request, a page view and a person are different units. A page can involve several files. One person can load the same page again. A click on a link is another event. The distinction determines what a report can mean; it is not a technical distraction from the business story.

The important opportunity was to observe certain reactions as part of the interaction itself. That did not automatically make human attention perfectly measurable. An image can be delivered without being consciously noticed. A link can be activated without a purchase following. The history of digital advertising includes the gradual clarification of these different objects of measurement.

It would also be misleading to suggest that advertising became measurable for the first time. Earlier media had research practices and ways of connecting advertisements with responses. What changed here was the close technical relationship between the placement, the link and a recordable movement. This affected the speed and organization of observation without creating the entire idea of effectiveness research from nothing.

The emergence of a number can itself influence decision-making. Once a response is recorded, managers can begin comparing and negotiating around it. The question becomes not only whether measurement is available, but which event has become convenient enough to dominate discussion. The banner’s history is partly a history of that shift in managerial attention.

The famous 44 per cent was not a universal benchmark

The reported click rate is memorable, which helps explain its prominence in later retellings. A more careful account treats it as a participant-reported result from an exceptional early case. Without complete raw records, a precise denominator and an agreed period, it cannot support a clean comparison with a later campaign. Its evidential status is not improved merely by repetition.

Audience composition alone can alter the comparison. Early visitors to a new technology publication need not resemble a later mass audience. Novelty may also motivate action. Someone exploring what a linked advertising image does is in a different situation from someone who encounters the format repeatedly while pursuing another task.

This does not make the reported response worthless. It changes the explanation. A high level of interaction might reflect curiosity, a suitable context or an attractive invitation. It is not a direct measure of trust, incremental revenue or customer retention. Each of those outcomes would require different evidence.

Nor should a later lower click rate be treated as a complete verdict on the quality of later creative work. Contexts, audiences, purposes and counting practices can differ. An apparent historical decline can contain several changes at once. A meaningful comparison would need to separate them rather than turning two percentages into a simple story of deterioration.

The case remains important without a miraculous performance claim. AT&T participated in an unusual publishing environment and connected its message to an exploratory experience. That arrangement has historical significance even when the remembered rate is presented cautiously. The explanation becomes stronger when it does not depend on treating every uncertainty as settled.

The web required more than transferring a print advertisement

Donna L. Hoffman and Thomas P. Novak’s 1996 Journal of Marketing article developed a conceptual account of marketing in networked hypermedia environments, incorporating navigation and flow. It was not an evaluation of the AT&T banner. It demonstrates that contemporary researchers were already asking how the medium’s structure changed the experience of communication (Hoffman and Novak, 1996).

For the banner case, this opens a useful perspective. Users could choose a route through connected documents. An advertisement therefore competed with the purposes guiding that navigation, not only with other advertisements. A creative execution needed to offer a possible next step that made sense in relation to the person’s activity.

The AT&T design made the transition particularly explicit. On the first page, it generated an invitation. Beyond the link, another environment had to provide a worthwhile continuation. Looking only at the visible strip misses part of the work. The small advertisement could open a much larger experience.

This does not imply that adding more interaction is always an improvement. An extra action can help someone or consume time without helping them. A technical capability is not identical to a user benefit. The relationship between message and activity must still be understood.

The historical significance is that media economics and user experience were connected from the outset. A publisher wanted to finance content, an advertiser wanted a valuable contact and a visitor wanted a worthwhile activity. The banner had to operate within all three relationships. It was never only a graphic object, even when the graphic itself was very simple.

A click and a change in brand perception were different outcomes

Rex Briggs and Nigel Hollis investigated whether web advertisements could have effects without click-through. Their 1997 Journal of Advertising Research article reported changes in advertising awareness and brand perceptions for the advertisements studied. It challenged the equation of not clicking with no effect, rather than retrospectively testing AT&T’s 1994 launch execution (Briggs and Hollis, 1997).

The distinction fits the broader problem. A person can notice a message without following its link. Another can click out of curiosity without developing a favourable opinion of the advertiser. Behaviour and attitude may be related, but one cannot simply stand in for the other.

The convenience of observing clicks could therefore encourage an overly narrow definition of success. What is easy to count can become the preferred measure, while other objectives become less visible in the reporting system. The objectives do not disappear. They require an appropriate way of investigating them.

The answer is neither to discard behavioural measures nor to relabel every disappointing result as branding. It is to identify the intended outcome before assessing it. A campaign designed to start an experience, provide information or improve recognition may raise different questions. The early research is important because it shows that these distinctions were being examined only a few years after HotWired’s debut.

Banner blindness concerned tasks as well as advertisements

Jan Panero Benway and David M. Lane’s 1998 Rice University research explored why users missed conspicuous banner-like elements, including links that could help with their assigned tasks. The phenomenon was therefore not restricted to deliberate rejection of paid advertising. Their work showed that making an element visually prominent did not ensure its use during purposeful navigation (Benway and Lane, 1998).

This complicates another familiar historical narrative. Reduced interaction need not be explained solely by worse design. People learn ways of scanning a page and identifying the information they need. A visually separate area may fall outside the route they expect to follow. Increasing its colour or size does not necessarily solve that problem.

The tension is particularly useful when interpreting an early experimental audience. An unfamiliar strip could itself become an object of exploration. Once users developed expectations about advertising areas, a similar appearance might help them bypass it. Audiences were changing participants in the medium, rather than a fixed population waiting to receive more sophisticated advertisements.

The study does not explain why a particular visitor clicked AT&T’s banner four years earlier. It provides an additional analytical perspective. A history of advertising should consider audience learning alongside changes in design and technology. Screen presence and mental attention are different conditions.

Publishers were questioning the economics by 1996

In a December 1996 WIRED article, Hunter Madsen identified HotWired as his company and discussed the financial difficulties of web publishers. He treated advertising as a central funding option while criticizing limitations of the banner. The source shows that concerns about the format’s economics and design appeared early within the publishing business (Madsen, 1996).

A celebrated advertisement and a viable publisher are different outcomes. Advertising revenue depends on prices, available placements, audiences and the costs of producing content. One memorable campaign does not prove that every publication selling similar strips can finance its operations.

Publishers also faced a tension between funding a service and preserving its usefulness. A placement can be conspicuous in the short term while making the overall experience less attractive. More commercial space does not necessarily produce a sustainable relationship with readers. That structural problem did not wait for later, more complex advertising technology.

Madsen’s article should be read as a contemporary intervention. Its expectations about future spending were forecasts, not results already achieved. That distinction is precisely what makes it valuable. It reveals how participants understood uncertainty before later market growth became part of the historical record.

An industry emerged around the individual experiments

On 25 March 1997, the IAB announced a reported 1996 internet advertising total of US$267 million through its programme administered by Coopers & Lybrand. The programme covered multiple forms of internet advertising. The figure is not worldwide banner revenue and cannot be assigned wholly to HotWired’s influence. It documents an early attempt to measure an emerging business collectively (Internet Advertising Bureau, 1997).

The form of the publication matters as well as the total. A developing industry needs common accounts of its scale. Sellers want to demonstrate relevance, while buyers need information for budget decisions. Organized research can help establish that conversation, provided its boundaries and institutional interests remain visible.

The comparison with 1994 reveals institutional development. The early problem was how to sell participation in something new. Within a few years, organizations were recommending dimensions, publishing research on effects and reporting aggregated revenues. Those arrangements made transactions more understandable and repeatable.

HotWired was an important location in this process, not its sole cause. Expansion required additional publishers, agencies, advertisers, technical suppliers and researchers. Recognizing their work preserves the significance of the pioneering example without turning the subsequent market into an automatic consequence of one creative execution.

A familiar funding principle acquired a different interface

Lilly Anne Buchwitz’s 2015 CHARM paper compared the development of radio and internet advertising through a periodization of technological, content and commercial change. Her framework helps place the banner within a process of negotiation over a medium’s economic form. Advertising support was familiar; an accepted implementation within the web still had to be developed (Buchwitz, 2015).

The AT&T example combined old and new relationships. A publisher offered an environment, a brand paid for presence and creative practitioners shaped a message. The message could now initiate an immediate transition, with the next experience forming part of what had been promised. The link extended the advertisement beyond its visible boundaries.

This also helps explain the case’s international significance. European web infrastructure, American publishing and agency work, and cultural destinations elsewhere belonged to the same constellation. Technical reach did not erase differences of language, access or practical usability. A worldwide medium did not instantly produce one uniform audience.

What changed on 27 October 1994

HotWired and AT&T represent a moment when graphical web advertising became a conspicuous commercial proposition for major brands. The advertisement invited action as well as observation. Its relationship with “You Will” and the cultural experience beyond the click gave a small image a wider purpose.

The detailed history is less tidy than the legend. Commercial web services came earlier, several advertisements accompanied the launch and explicit industry size recommendations followed later. The celebrated click rate is a remembered participant figure. Neither the effectiveness of advertising nor the economics of publishing can be reduced to that one number.

Those qualifications reveal the achievement more clearly. The banner became part of an arrangement connecting publisher, brand, creative work, destination and measurement. AT&T’s execution remains the most recognizable early symbol because it made the unfamiliar action its subject. A picture was no longer only something to look at. It could become the beginning of another experience.

References

Benway, J.P. and Lane, D.M. (1998) ‘Banner Blindness: Web Searchers Often Miss “Obvious” Links’, Rice University, research report. https://www.ruf.rice.edu/~lane/papers/banner_blindness.pdf.

Briggs, R. and Hollis, N. (1997) ‘Advertising on the Web: Is There Response before Click-Through?’, Journal of Advertising Research, 37(2), pp. 33–45. https://doi.org/10.1080/00218499.1997.12466659.

Buchwitz, L.A. (2015) ‘A Model of Periodization of Radio and Internet Advertising History’, Proceedings of the Conference on Historical Analysis and Research in Marketing, 17, pp. 14–29. https://ojs.library.carleton.ca/index.php/pcharm/article/view/1465.

CERN (n.d.) ‘Licensing the Web’. https://home.cern/science/computing/the-birth-of-the-web/licensing-web/.

Computer History Museum (n.d.) ‘Surfing the Web in the Early 1990s’, Revolution. https://www.computerhistory.org/revolution/the-web/20/408.

Digiday (2013) ‘How the Banner Ad Was Born’, 12 April. Interview-based reconstruction. https://digiday.com/marketing/how-the-banner-ad-was-born/.

Digiday (2014) ‘Banner ad creator: “My kids make fun of me for it”’, 2 April. Interview with Joe McCambley. https://digiday.com/media/banner-ad-creator/.

Digiday (2017) ‘“The beginning of a giant industry”: An oral history of the first banner ad’, 8 November. https://digiday.com/media/history-of-the-banner-ad/.

Hoffman, D.L. and Novak, T.P. (1996) ‘Marketing in Hypermedia Computer-Mediated Environments: Conceptual Foundations’, Journal of Marketing, 60(3), pp. 50–68. https://doi.org/10.1177/002224299606000304.

Internet Advertising Bureau and CASIE (1996) ‘Voluntary Guidelines for Banner Advertising and Process for Exploring Future Internet Advertising Forms Announced by Industry Groups’, 10 December. https://www.iab.com/news/voluntary-guidelines-banner-advertising-process-exploring-future-internet-advertising-forms-announced-industry-groups/.

Internet Advertising Bureau (1997) ‘Internet Advertising Bureau Announces 1996 Advertising Revenue Reporting Program Results’, 25 March. https://www.iab.com/news/internet-advertising-bureau-announces-1996-advertising-revenue-reporting-program-results/.

Madsen, H. (1996) ‘Reclaim the Deadzone’, WIRED, 1 December. https://www.wired.com/1996/12/esmadsen/.

O’Reilly Media (n.d.) ‘History from 1984 to the Present’. https://www.oreilly.com/about/history.

Web Design Museum (n.d.) ‘The first web banner’. https://www.webdesignmuseum.org/web-design-history/the-first-web-banner-1994.

Few psychological ideas have entered marketing culture as completely as Maslow’s hierarchy of needs. It appears in brand workshops, consumer-behaviour textbooks, advertising presentations and segmentation exercises across the world. Its familiar pyramid seems to offer an elegant explanation of human motivation: people first seek food and physical survival, then safety, belonging, esteem and, finally, self-actualisation.

For marketers, the appeal is obvious. Insurance can be positioned around safety, social platforms around belonging, luxury goods around esteem and education or travel around self-actualisation. The model translates abstract human motivation into a simple set of communication territories.

Yet the famous pyramid is not an accurate reproduction of Abraham Maslow’s original theory. Maslow never drew a pyramid. His 1943 article “A Theory of Human Motivation” described five broad sets of goals, but treated their order as relative rather than mechanically fixed. Needs could be partially satisfied, several could operate simultaneously, and individual priorities could differ (Maslow, 1943).

The triangular image emerged later, when management writers and consultants converted Maslow’s theory into a compact tool for organisational motivation. Bridgman, Cummings and Ballard traced the now-familiar pyramid to Charles McDermid’s 1960 management article, rather than to Maslow himself (Bridgman, Cummings and Ballard, 2019).

This difference matters. Maslow developed a humanistic account of motivation and growth. The pyramid turned that account into a visual technology of management.

Marketing inherited both versions. From Maslow it gained the idea that consumers pursue psychological, social and developmental goals beyond simple survival. From the pyramid it gained an apparently universal ordering system through which products and messages could be classified.

The first contribution remains valuable. The second requires caution.

The history of Maslow’s hierarchy therefore reveals a broader pattern in marketing thought. Complex theories are often simplified when they enter business education. The simplified version may become more influential than the original, even when empirical research offers limited support.

A historically informed approach does not require marketers to abandon Maslow. It requires them to use his theory as a flexible heuristic rather than as a scientifically proven map of every consumer.

Abraham Maslow and Humanistic Psychology

Abraham Harold Maslow was born in 1908 and became a central figure in the development of humanistic psychology. He taught at Brooklyn College and later at Brandeis University, where he worked from 1951 to 1969 and helped establish its psychology department.

Maslow’s work developed partly in response to two dominant psychological traditions. Psychoanalysis concentrated heavily on unconscious drives, conflict and pathology. Behaviourism explained behaviour through observable learning, reinforcement and environmental conditioning.

Maslow believed that these approaches illuminated important aspects of human life but neglected creativity, purpose, personal growth and psychological health.

Humanistic psychology therefore shifted attention from dysfunction alone towards human possibility. Rather than asking only what causes neurosis or maladaptive behaviour, Maslow asked what enables people to flourish.

This perspective later appealed strongly to marketers. A theory centred on aspiration and personal development could explain why consumers seek experiences, identities and meanings rather than only material utility.

Maslow’s influence extended well beyond psychology into management, education, healthcare, design and consumer research. Ironically, the simplified pyramid often travelled further than his original writings.

The Original 1943 Theory

Maslow introduced his theory in the Psychological Review in 1943. He proposed at least five sets of basic goals: physiological needs, safety, love and belonging, esteem and self-actualisation (Maslow, 1943).

Physiological needs include food, water, sleep and other bodily requirements. Maslow argued that severe deprivation can dominate consciousness. A starving person may organise thought and behaviour around the acquisition of food.

When physiological needs are sufficiently satisfied, safety can become more prominent. Safety concerns protection, stability, order and freedom from threat.

Love and belonging involve affectionate relationships, friendship, family and participation in groups.

Esteem includes both internal and external dimensions. People seek competence, mastery, independence and self-respect, but also recognition, reputation and status.

Self-actualisation refers to the desire to realise one’s potential. Maslow argued that a person must become what he or she is capable of becoming. The specific form differs: one person may realise potential through art, another through science, parenting, public service or business.

Later textbook accounts transformed these categories into five separate floors of a pyramid. Maslow’s original article did not.

Relative Priority Rather than Absolute Sequence

The popular rule that one level must be completed before the next begins is a distortion.

Maslow described an average pattern of relative prepotency. A stronger unmet need tends to organise behaviour, but needs do not switch on and off like lights.

He explicitly suggested that an individual might be satisfied to different degrees across all levels simultaneously. He offered an illustrative example in which physiological needs might be 85 per cent satisfied, safety 70 per cent, love 50 per cent, esteem 40 per cent and self-actualisation 10 per cent satisfied (Maslow, 1943).

He also acknowledged reversals and exceptions. Some individuals may value esteem above love. Creativity may remain compelling despite poverty. Ideals, faith or political commitments may lead people to sacrifice safety.

Behaviour is also multiply motivated. A meal may satisfy hunger, reinforce family bonds, express religious identity and communicate social status.

This insight is essential for marketing. Products cannot be placed objectively on a single level. Their meaning changes across consumers and occasions.

A smartphone can provide emergency security, social belonging, prestige and creative expression. A university degree can promise employment security, group membership, status and intellectual growth.

The hierarchy is therefore better understood as a set of interacting motivational domains than as a rigid staircase.

Maslow Did Not Create the Pyramid

The absence of a pyramid in Maslow’s work is more than a historical curiosity. Visual form changes interpretation.

Bridgman, Cummings and Ballard investigated the evolution of the image and concluded that Maslow never used it. Management educators began presenting the hierarchy through simplified stages during the 1950s. Keith Davis used a stepped triangular illustration, and Charles McDermid published a recognisable pyramid in 1960 (Bridgman, Cummings and Ballard, 2019).

McDermid connected the hierarchy to managerial efficiency and employee motivation. The model helped managers imagine that workers could be moved from wages and security towards recognition and achievement.

The pyramid’s broad base and narrow peak also implied that lower needs were common while self-actualisation was rare and elite.

This visual simplification fitted post-war management education. It was memorable, easy to reproduce and compatible with a managerial desire to organise complex human behaviour.

Once included in textbooks, the image became self-reinforcing. Students encountered the pyramid, later used it as managers and consultants, and taught it to others.

Marketing adopted the same diagram because it offered a convenient way to organise consumer motives.

Why Marketing Adopted Maslow

Marketing seeks to understand why people value one offering over another. Functional product attributes are only part of the answer.

Consumers purchase food, but also comfort and ritual. They purchase clothing, but also identity and social acceptance. They purchase transport, but also control, freedom and status.

Maslow gave marketers an accessible vocabulary for such differences.

Kotler and Keller distinguish needs, wants and demands. Needs are fundamental human requirements. Wants are culturally and personally shaped forms through which needs are expressed. Demands are wants supported by purchasing power (Kotler and Keller, 2016).

Maslow’s categories appeared to organise the first part of this process. Marketers could ask which need an offering serves and then investigate how culture turns that need into specific wants.

The framework was particularly suited to positioning. A brand could move beyond product features and articulate an emotional or social benefit.

A home-security service is technically a network of sensors, but its value proposition is protection and peace of mind. A premium watch measures time, yet may communicate achievement, craftsmanship and continuity.

Maslow’s influence therefore supported the broader historical movement from product-centred selling towards benefit-, identity- and value-based marketing.

Physiological Needs and Consumer Markets

Physiological needs are associated with food, drink, shelter, sleep and bodily comfort.

Many large consumer markets originate in these requirements. Food manufacturers, restaurants, beverage brands, housing providers, clothing companies and hospitality businesses all address basic bodily conditions.

However, modern competition rarely remains at the level of need satisfaction alone. Numerous products can relieve hunger or thirst. Differentiation occurs through taste, convenience, health, ethics, identity and experience.

Coffee may provide stimulation, but it can also signify routine, sophistication or social connection. A hotel room provides shelter and sleep, while a premium hotel sells personal attention, status and escape.

Even a basic need is culturally mediated. People do not consume nutrients in the abstract. They consume cuisines, brands, rituals and socially meaningful occasions.

For marketers, the physiological category should therefore be a starting point, not a complete explanation.

Safety as a Marketing Proposition

Safety marketing is widespread because threat and uncertainty are powerful motivators.

Insurance, financial services, medicine, cybersecurity, automobiles and home technology frequently emphasise protection and predictability.

Trust signals are part of the same domain. Warranties, independent testing, certifications, secure payment systems and transparent returns reduce perceived risk.

Automobile campaigns may show crash performance and driver-assistance systems. Banks communicate stability and fraud protection. Medical companies stress clinical evidence and quality controls.

The ethical risk is fear amplification. A campaign can inform consumers about genuine risk, but it can also exaggerate danger to create anxiety.

Fear appeals are especially sensitive in healthcare, finance and family protection. Responsible marketing should provide evidence, proportionality and meaningful choices rather than implying that refusal to buy is reckless.

Belonging and Brand Communities

The desire for belonging is central to advertising and branding.

Family meals, friendship groups, sports teams and shared celebrations appear repeatedly in commercial communication. They connect products with relationships rather than isolated consumption.

Brands can also become social markers. Fans of a sports club, motorcycle brand, music genre or game may experience real community through shared symbols and rituals.

Social media made belonging measurable and continuously accessible. Following an account, joining a group or using a hashtag can signal affiliation.

This creates value for brands, but also responsibility. Community marketing can foster genuine participation. It can equally exploit fear of exclusion.

Scarcity, invitation-only access and “everyone is joining” messages can make consumers feel socially inadequate.

The useful question is not merely whether belonging drives engagement, but whether a brand contributes to meaningful connection or manufactures insecurity.

Esteem, Recognition and Status Consumption

Esteem covers competence, achievement, reputation and recognition.

Luxury branding frequently addresses visible status. Watches, cars, fashion and travel may serve as social signals.

Yet esteem is broader than prestige. It also includes mastery, autonomy and self-respect.

Bourdieu demonstrated that consumption and taste help reproduce social distinction (Bourdieu, 1984). Levy argued that goods function as symbols through which people communicate identities and relationships (Levy, 1959).

Marketing can therefore appeal to external recognition or internal competence.

A professional qualification may promise promotion and status, but also genuine knowledge. A sports product may suggest social prestige, but also progress and personal mastery.

The distinction matters strategically. Messages built around superiority differ from those built around capability.

It also matters ethically. Marketing that repeatedly tells people they are inadequate without a particular product may intensify status anxiety.

Self-Actualisation and Aspirational Marketing

Self-actualisation has become the most attractive category for aspirational brands.

Educational institutions, travel companies, creative software providers, wellness services and coaching businesses often promise personal development.

The consumer is shown a possible future self: more capable, creative, independent or fulfilled.

This form of marketing can help people imagine meaningful change. It can also commercialise dissatisfaction.

When every hobby becomes self-optimisation and every experience becomes an investment in personal branding, consumers may feel permanently incomplete.

Self-actualisation should therefore not be reduced to continuous achievement. Maslow associated it with authenticity and the realisation of individual potential, not with endless consumption.

A product can support development, but cannot guarantee a meaningful life. Marketing claims should respect this limit.

Multi-Level Positioning

The strongest practical use of Maslow lies in recognising that one offering can address several needs.

A fitness platform can support physical health, reduce health risk, create community, provide achievement and enable personal growth.

A university can promise employment security, friendship networks, status and intellectual development.

A family car can provide transport, physical safety, social belonging and recognition.

This multi-level analysis is more realistic than assigning each category to one box.

It also reveals why different segments respond to different messages. One consumer may choose a vehicle primarily for reliability; another for family protection; another for design and prestige.

Marketing research must identify which motives are active rather than assuming them from the product category.

Social Media through a Maslovian Lens

Social platforms combine belonging, esteem and self-expression in unusually visible ways.

Users maintain relationships, join communities and share interests. At the same time, likes, follower counts and views quantify recognition.

Creators can use platforms for artistic expression and entrepreneurship, suggesting self-actualisation. Yet their visibility depends partly on algorithmic systems and public approval.

Influencer marketing operates across several layers. Followers may feel social closeness to a creator, admire the creator’s status and purchase products associated with that lifestyle.

Brands can participate in these relationships through sponsorships and community content.

The model also exposes platform risks. If recognition is quantified, users may become dependent on feedback. Idealised images can intensify comparison. Fear of missing out can turn belonging into pressure.

For marketers, Maslow should function here as an ethical audit. Which need is being addressed? Does the campaign support it or deliberately destabilise it?

The Evidence Problem

Maslow’s hierarchy became enormously popular without receiving equally strong empirical support.

Wahba and Bridwell reviewed research on the theory in 1976. They found little consistent evidence for a strict five-level ordering and limited support for the proposition that satisfaction of one need automatically activates the next (Wahba and Bridwell, 1976).

The categories themselves remain intuitively meaningful. The difficulty lies in their fixed order.

People maintain love, creativity and meaning under conditions of poverty and insecurity. Others with substantial material security may continue to focus primarily on safety.

Motivation changes with personality, age, life events and social context.

Tay and Diener studied need fulfilment and well-being across 123 countries. Basic, social and respect-related needs were all associated with well-being, but higher needs remained relevant when lower needs were not fully satisfied (Tay and Diener, 2011).

The evidence supports the importance of multiple human needs more clearly than it supports the pyramid’s sequencing.

Cultural Bias

Maslow’s model has often been criticised as culturally individualistic.

Hofstede argued that placing personal self-actualisation at the summit reflects mid-twentieth-century American values rather than a neutral universal ordering (Hofstede, 1984).

In collectivist contexts, family duty, communal harmony and social belonging may not be lower stages on the way to individual fulfilment. They may constitute fulfilment itself.

International marketers must therefore avoid mapping national cultures onto a universal pyramid.

A campaign celebrating independence and personal choice may work in one context and appear selfish in another. Status may be expressed through individual luxury, family success, educational achievement or community contribution.

The underlying needs may recur, but their meanings and priorities are culturally produced.

The Blackfoot and Siksika Debate

Maslow spent time with the Siksika community in Alberta in 1938. This encounter is sometimes presented online as the hidden source of his hierarchy.

Some accounts claim that Maslow borrowed an Indigenous model that placed community and cultural continuity above individual self-actualisation.

The historical evidence does not support a simple claim that he copied a pre-existing Blackfoot pyramid. No documented Indigenous pyramid identical to the later management diagram has been established.

However, First Nations scholars have argued that Maslow’s exposure to Siksika social life may have influenced his thinking about security, cooperation and human potential. They also stress that Indigenous understandings of well-being are relational and intergenerational and should not be reduced to the Western pyramid (Blackstock et al., 2022).

A careful history should acknowledge the visit and its possible influence without replacing one simplified origin myth with another.

Methodological Criticism

Maslow’s account of self-actualising individuals was not based on a representative population sample.

He selected people he considered psychologically healthy and admirable, including historical figures and individuals known to him. The criteria reflected his own judgement.

This creates a methodological problem. If self-actualisation is defined through selected characteristics and people are then chosen because they display those characteristics, the argument risks circularity.

The examples also overrepresented educated, Western and socially successful individuals.

A wider understanding of human fulfilment must recognise care work, communal responsibility, craft, spirituality and other forms of achievement that may not resemble elite intellectual or creative careers.

Marketing should similarly avoid assuming that every consumer’s highest aspiration is entrepreneurial, individualistic or publicly visible.

Deficiency and Growth Motivation

Maslow later distinguished deficiency needs from growth motivation.

Deficiency needs arise through lack. Hunger, danger, loneliness or humiliation create pressure to restore an acceptable condition.

Growth motivation operates differently. Learning and creativity may generate further interest rather than eliminate it.

This distinction remains useful for positioning.

Some products solve an acute problem. Others enable continuing development. A security service reduces threat; an educational service may expand curiosity.

Marketers should identify whether a proposition offers relief, maintenance or growth.

They should also avoid presenting every purchase as transformational. Not every ordinary product needs a self-actualisation narrative.

Later Extensions and Transcendence

Maslow’s later work included cognitive, aesthetic and transcendence-related motivations.

Cognitive needs concern knowledge and understanding. Aesthetic needs concern order, beauty and form. Transcendence involves commitment to values or purposes beyond the individual self.

These additions are relevant to cultural institutions, education and purpose-led organisations.

They also demonstrate that the famous five-level pyramid freezes one stage in Maslow’s evolving thought.

Purpose marketing can appeal to transcendence, but only when organisational behaviour supports the claimed values. A brand cannot credibly promise environmental or social contribution while its operations contradict the message.

Evolutionary Revisions

Kenrick and colleagues proposed a renovated hierarchy based on evolutionary psychology. Their model retained immediate physiological and safety concerns but added motives such as mate acquisition, mate retention and parenting (Kenrick et al., 2010).

They treated motivational systems as overlapping rather than permanently replaced.

The revised model stimulated debate because it removed self-actualisation from the summit and emphasised reproductive motives.

Its significance lies less in providing a final replacement than in demonstrating that Maslow’s pyramid is not the only possible hierarchy.

Motivational models depend on theoretical assumptions about what human beings are and what goals count as fundamental.

Alternatives for Marketing Research

Maslow should be combined with more specific theories.

Self-determination theory identifies autonomy, competence and relatedness as important psychological needs (Deci and Ryan, 2000). It offers a strong basis for analysing engagement, loyalty and intrinsic motivation.

Means–end chain theory connects product attributes with consequences and personal values. Identity theories explain symbolic consumption. Behavioural economics examines biases and decision contexts. Jobs-to-be-Done focuses on the progress consumers seek in particular situations.

These approaches often generate more testable marketing hypotheses.

Maslow remains valuable because it stimulates broad questions, not because it supplies precise predictions.

Ethical Use in Marketing

Needs-based marketing can become manipulative when it intentionally creates or deepens vulnerability.

A beauty brand may imply that social acceptance depends on appearance. A financial provider may exaggerate insecurity. A platform may design recognition metrics that encourage compulsive engagement.

Children, patients and financially vulnerable consumers require particular care.

The ethical standard should move beyond legal disclosure. Marketers should consider whether a campaign respects autonomy and provides truthful information.

Berghoff’s concept of marketing as a modern social technology is relevant here. Marketing does not merely discover natural demand. It helps organise markets, meanings and consumer expectations (Berghoff, 2007).

Using psychological insight therefore carries social responsibility.

A Better Way to Use Maslow

The model works best as a flexible set of questions.

What physical or functional problem does the offering address? What uncertainty does it reduce? Which relationships does it enable? What forms of competence or recognition does it support? How might it contribute to development, meaning or creativity?

These questions should be answered through consumer research rather than assumption.

Marketers should expect several needs to interact. They should investigate cultural variation and distinguish genuine product value from promotional fantasy.

Maslow can begin analysis. It should not end it.

Conclusion

Maslow’s hierarchy of needs occupies an unusual position in marketing history. It is one of the field’s most recognisable psychological models, yet the version most frequently taught was not created by Maslow.

His 1943 theory described five broad categories of motivation and proposed a relative ordering. It allowed partial satisfaction, overlapping motives and exceptions.

The pyramid was constructed later by management educators and consultants who wanted a clearer tool for organisational practice. Its visual simplicity drove its international spread.

Marketing adopted the model because it helped move attention from product attributes towards human motives. Brands could position themselves around safety, belonging, esteem and growth.

This contribution remains important. Consumers do not purchase only material objects. They purchase protection, connection, identity, competence and imagined futures.

But the model’s limitations are equally important. Evidence does not support a rigid universal hierarchy. Cultural contexts shape the meaning and priority of needs. Maslow’s samples and methods were limited. The famous pyramid conceals his later revisions and the complexity of his original account.

The responsible conclusion is neither to worship nor discard Maslow.

His theory should be treated as a historical heuristic: a useful prompt for considering multiple dimensions of value, but not a universal law of consumer behaviour.

Modern marketing requires a more flexible image than a pyramid. Human motivation resembles an interacting network in which physical, social, cultural and personal goals change across situations.

The task of marketing is not to place every person on the correct step. It is to understand those changing relationships and create genuine value without exploiting the vulnerabilities they reveal.

References

Berghoff, H. (ed.) (2007) Marketinggeschichte: Die Genese einer modernen Sozialtechnik. Frankfurt am Main and New York: Campus.

Blackstock, C., Cross, T., George, J. and Brown, I. (2022) ‘Reconsidering Maslow and the hierarchy of needs from a First Nations’ perspective’, Aotearoa New Zealand Social Work, 34(2), pp. 30–41.

Bourdieu, P. (1984) Distinction: A Social Critique of the Judgement of Taste. Cambridge, MA: Harvard University Press.

Bridgman, T., Cummings, S. and Ballard, J. (2019) ‘Who built Maslow’s pyramid? A history of the creation of management studies’ most famous symbol and its implications for management education’, Academy of Management Learning & Education, 18(1), pp. 81–98.

Deci, E.L. and Ryan, R.M. (2000) ‘The “what” and “why” of goal pursuits: Human needs and the self-determination of behavior’, Psychological Inquiry, 11(4), pp. 227–268.

Hofstede, G. (1984) ‘The cultural relativity of the quality of life concept’, Academy of Management Review, 9(3), pp. 389–398.

Kenrick, D.T., Griskevicius, V., Neuberg, S.L. and Schaller, M. (2010) ‘Renovating the pyramid of needs: Contemporary extensions built upon ancient foundations’, Perspectives on Psychological Science, 5(3), pp. 292–314.

Kotler, P. and Keller, K.L. (2016) Marketing Management. 15th edn. Harlow: Pearson.

Levy, S.J. (1959) ‘Symbols for sale’, Harvard Business Review, 37(4), pp. 117–124.

Maslow, A.H. (1943) ‘A theory of human motivation’, Psychological Review, 50(4), pp. 370–396.

Maslow, A.H. (1954) Motivation and Personality. New York: Harper & Brothers.

Tay, L. and Diener, E. (2011) ‘Needs and subjective well-being around the world’, Journal of Personality and Social Psychology, 101(2), pp. 354–365.

Wahba, M.A. and Bridwell, L.G. (1976) ‘Maslow reconsidered: A review of research on the need hierarchy theory’, Organizational Behavior and Human Performance, 15(2), pp. 212–240.

Few marketing concepts have travelled as successfully as four short words beginning with the same letter. Product, Price, Place and Promotion appear in introductory textbooks, business plans, university examinations, consultancy presentations and corporate strategy workshops around the world. They are often presented as if they had always belonged together, as though one scholar had simply discovered four timeless laws of marketing and placed them in a memorable diagram.

The actual history is more complex. The 4Ps were not created in a single intellectual moment, and they were not invented by Philip Kotler, despite the frequent association of his name with the framework. Their development resulted from several decades of debate about what marketing managers controlled, how marketing decisions should be classified and how an emerging academic discipline could be transformed into a practical management system.

James Culliton supplied the metaphor that inspired the concept. Neil H. Borden developed and popularised the expression “marketing mix”. E. Jerome McCarthy condensed a much larger collection of marketing variables into the four categories now known as Product, Price, Place and Promotion. Philip Kotler subsequently integrated the model into an influential system of marketing analysis, planning and control, helping generations of students and executives adopt the framework.

The history of the marketing mix is therefore also the history of marketing’s transformation. During the early twentieth century, marketing scholarship was strongly concerned with commodities, institutions and the functions involved in moving goods from producers to consumers. By the middle of the century, attention increasingly shifted towards the decisions of the marketing manager. The 4Ps expressed this managerial turn. Marketing was no longer treated only as a collection of market activities. It became a coordinated set of variables that a company could deliberately combine in pursuit of a target-market response.

This shift had far-reaching consequences. The framework made marketing easier to teach, communicate and apply. It encouraged managers to think beyond advertising and recognise that product design, pricing and distribution were also marketing decisions. At the same time, its simplicity created limitations. Critics argued that the 4Ps reflected the perspective of the seller rather than the customer, fitted manufactured goods better than services and could make marketing appear as a collection of controllable tactical levers rather than a relational, institutional and social process.

Digital platforms and social media have intensified these debates. On Instagram, TikTok, YouTube and other platforms, a product may be developed together with a creator community, its price may change dynamically, its distribution may occur through a social-commerce interface and its promotion may be inseparable from entertainment or personal communication. The four categories remain recognisable, but the boundaries among them have become increasingly porous.

The history of the 4Ps therefore matters for more than academic accuracy. Understanding where the model came from reveals what it was originally designed to do, why it became so influential and where it can mislead modern marketers. The marketing mix remains useful, but only when it is treated as a decision framework rather than an unquestionable definition of marketing itself.

Marketing before the Marketing Mix

Marketing practices are much older than the academic discipline that studies them. Merchants, craftspeople and market institutions have long made decisions about products, prices, locations, distribution and persuasive communication. Ancient and medieval societies contained marketplaces, maker’s marks, retail signs, salespeople, transport networks and commercial messages. Eric H. Shaw’s analysis of ancient and medieval marketing demonstrates that exchange-related practices should not be confused with the much later emergence of formal marketing thought (Shaw, 2016).

The modern academic field developed primarily during the late nineteenth and early twentieth centuries as industrialisation, urbanisation, national distribution and mass production created new problems of market coordination. Manufacturers produced goods in greater volumes and sold them across longer distances. Wholesalers, retailers, transport companies, advertising agencies and market researchers became increasingly important.

Early marketing scholars did not initially organise the field around the decisions of a single brand manager. They commonly studied commodities, institutions and functions.

The commodity approach examined how particular goods, such as grain, cotton or manufactured products, moved through markets. The institutional approach focused on organisations such as wholesalers, retailers, brokers and exchanges. The functional approach classified the activities required to complete marketing processes, including buying, selling, transportation, storage, financing, risk bearing and market information.

Robert Bartels’ history of marketing thought shows how these different schools contributed to the formation of marketing as a recognisable academic discipline (Bartels, 1988). D. G. Brian Jones and Mark Tadajewski likewise emphasise that marketing theory evolved through changing economic, institutional and intellectual contexts rather than following one inevitable path towards modern management (Jones and Tadajewski, 2016).

The 4Ps emerged when the managerial approach began to displace or reorganise these earlier perspectives. Instead of asking only which functions existed in a market system, scholars increasingly asked which decisions a marketing executive had to make.

This distinction was fundamental. A functional description might explain that distribution, storage and selling take place. A managerial framework asks how a particular organisation should configure these activities to achieve its objectives.

James Culliton and the “Mixer of Ingredients”

The conceptual ancestry of the marketing mix is usually traced to James W. Culliton, a Harvard professor whose 1948 research bulletin was titled The Management of Marketing Costs. Culliton did not formulate the four Ps. Nor did he present a formal marketing-mix diagram. His contribution was a metaphor.

Culliton described marketing executives as people who combined different ingredients. Some followed established recipes, some adapted recipes to available resources, and others experimented with new combinations. The marketing manager was therefore a “mixer of ingredients”.

This metaphor captured the complexity of managerial decision-making. There was no single universally correct marketing programme. Managers combined variables according to products, competition, customer behaviour, distribution structures and organisational resources.

The culinary comparison was especially effective because it emphasised interaction. Ingredients do not operate independently. Changing one element may alter the effect of the whole mixture. A lower price might support mass distribution but undermine an exclusive brand position. Premium packaging could reinforce quality perceptions but increase cost. Intensive advertising could generate demand that an inadequate distribution system could not satisfy.

The later 4P model inherited this central insight. The Ps were not intended as four separate checkboxes. They formed a coordinated mix.

Neil Borden later explained that Culliton’s description of the marketing executive as a mixer of ingredients inspired him to use the term “marketing mix”. Borden began employing the phrase in his teaching and writing during the late 1940s and used it publicly before his famous 1964 article (Borden, 1964). Historical research presented through the CHARM tradition has likewise traced the transition from Culliton’s metaphor to Borden’s concept and McCarthy’s later classification (Silverman, 1995).

Neil H. Borden and the Creation of the Marketing-Mix Concept

Neil Hopper Borden was a professor of advertising at Harvard Business School and a former president of the American Marketing Association. His career connected advertising research with the growing managerial orientation of marketing.

Borden did not view the marketing mix as four variables. His framework was much more extensive. In his 1964 article, The Concept of the Marketing Mix, he described twelve groups of marketing ingredients:

product planning, pricing, branding, channels of distribution, personal selling, advertising, promotions, packaging, display, servicing, physical handling, and fact finding and analysis.

This list reflected the complexity of post-war marketing management. A marketing executive had to make decisions about what to sell, what name and package to use, how much to charge, where the product should be available, how it should be advertised, which salespeople should represent it and how information should be collected.

Borden also identified market forces that shaped the manager’s choices, including consumer behaviour, trade behaviour, competitor actions and government regulation. The marketing executive could combine controllable ingredients, but did so within an environment that could not be controlled completely.

This point is often lost in simplified explanations. Borden’s marketing mix was not a claim that companies possessed unlimited power over markets. It was a managerial response to uncertainty. The manager had to observe external forces and design a suitable combination of actions.

Borden regarded marketing-mix construction partly as an art. Reliable facts and systematic analysis were important, but management also required judgement. Competitors, customers and institutions responded dynamically, making a universally valid formula impossible.

The difference between Borden’s framework and the later 4Ps is therefore substantial. Borden provided an open-ended inventory and an image of the manager as a creative mixer. McCarthy supplied a compact taxonomy.

Borden is best described as the originator and principal early populariser of the marketing-mix concept, not the inventor of the 4P classification.

Other Attempts to Classify Marketing Decisions

The 4Ps did not emerge in an intellectual vacuum. During the 1950s and early 1960s, several scholars attempted to classify marketing variables more systematically.

John A. Howard organised managerial decisions around categories resembling product, distribution, communication and price. Albert W. Frey distinguished between the “offering” and the “methods and tools” used to market it. The offering included elements such as product, packaging, brand, price and service, while the methods included distribution, personal selling, advertising, sales promotion and publicity.

William Lazer and Eugene Kelley proposed another arrangement involving a goods-and-services mix, a distribution mix and a communication mix. These classifications demonstrate that scholars broadly agreed on the need to organise managerial decisions but did not yet agree on one definitive structure.

The eventual dominance of McCarthy’s system did not necessarily prove that it was theoretically superior in every respect. Its success also reflected clarity, memorability and educational usefulness.

Marketing concepts compete not only through empirical accuracy but through their ability to be taught and remembered. The alliterative structure of four Ps gave McCarthy’s framework a powerful advantage.

E. Jerome McCarthy and the Birth of the 4Ps

Edmund Jerome McCarthy published the first edition of Basic Marketing: A Managerial Approach in 1960. The book organised the marketing manager’s major controllable variables into four groups:

Product, Price, Place and Promotion.

McCarthy did not claim to have invented all the decisions contained within these categories. Product development, pricing, distribution and communication had long been recognised as important. His innovation was taxonomic and pedagogical: he grouped a wide range of marketing activities into four memorable managerial categories.

The original 1960 edition comprised approximately 770 pages and presented marketing as a broad managerial discipline rather than merely a four-part formula (McCarthy, 1960). The book covered customers, market analysis, segmentation, planning and environmental conditions alongside the marketing mix.

McCarthy’s four Ps can be understood as a compression of Borden’s larger list.

Product absorbed decisions concerning product planning, branding, packaging and many aspects of service. Price covered pricing policies, discounts and related financial terms. Place represented channels, physical distribution, logistics and market coverage. Promotion combined advertising, personal selling, sales promotion and publicity-related communication.

This condensation made the framework manageable. A student could remember four categories and then place more detailed decisions underneath them. An executive could use the categories to review a marketing plan.

McCarthy’s contribution should therefore be stated precisely:

James Culliton inspired the mixing metaphor. Neil Borden developed and popularised the marketing-mix concept. E. Jerome McCarthy created the four-P classification in 1960.

Product: More Than a Physical Object

The first P is frequently reduced to the item a company manufactures. McCarthy’s managerial approach was broader. Product decisions include the complete market offering: quality, features, design, assortment, brand name, packaging, guarantees, services and lifecycle management.

A product may be a physical good, a service, an experience, an organisation, a person, a place or an idea. Philip Kotler later played an important role in broadening marketing beyond commercial packaged goods, arguing that marketing principles could be applied to services, organisations, causes and public institutions.

Historically, Product brought product planning into marketing rather than leaving it solely to engineering or manufacturing. This was a significant managerial claim. A company should not produce something first and ask marketing merely to advertise it afterwards. Customer needs and market positioning should influence what is offered.

The product category also includes branding. A chemically similar item may acquire a different market position through its name, design, package, service and symbolic associations. David Aaker, Kevin Lane Keller, Jean-Noël Kapferer and Douglas Holt later developed sophisticated theories of brand identity, brand equity and cultural branding, but the Product element already recognised that an offering consisted of more than physical functionality (Aaker, 1991; Keller, 1993; Kapferer, 2012; Holt, 2004).

For social-media marketing, Product has become increasingly participatory. Companies monitor comments, creator content, reviews and community discussions when developing new features or variants. Software products can be changed after launch. Creators may collaborate in product development. Digital communities sometimes function as testing environments.

Social media therefore does not make Product irrelevant. It makes product decisions more visible, continuous and socially negotiated.

Price: Economic Exchange and Symbolic Positioning

Price is the only traditional P that directly generates revenue; the others generally require expenditure or investment. Yet pricing is not merely a calculation of cost plus margin.

Price communicates position. A high price may signal exclusivity or superior quality. A low price may support accessibility, market penetration or value positioning. Discounts, payment conditions, subscriptions, bundles and credit terms all influence the customer’s perceived exchange.

The inclusion of Price within marketing was important because it linked market strategy with financial decisions. A promotion can attract attention, but the final response depends partly on whether customers perceive the exchange as worthwhile.

Pricing interacts closely with the other Ps. A luxury product sold through exclusive distribution and refined communication requires a coherent price. Heavy discounting may attract short-term demand while weakening a premium identity.

Digital platforms have added new complexity. Prices can be tested, personalised or changed rapidly. Subscription models, freemium products, in-app purchases and creator memberships differ from conventional unit pricing.

Social-media platforms themselves demonstrate an unusual pricing structure. Users generally pay no monetary fee for access, while advertisers finance the system. The user receives a service but also contributes attention, behavioural data and content. Price must therefore be understood as part of a broader exchange rather than only the amount printed on a label.

This supports the later customer-oriented argument that marketers should consider total cost, including time, effort, risk, data and inconvenience.

Place: Distribution, Availability and Market Access

The word Place is perhaps the most misunderstood of the four Ps. It does not simply mean the physical location of a store. It includes all decisions involved in making an offering available to the intended customer.

These decisions include distribution channels, wholesalers, retailers, logistics, storage, transportation, inventory and market coverage. Robert D. Tamilia’s historical work on channels of distribution shows that marketing development cannot be separated from the institutions that connect production and consumption (Tamilia, 2016).

Place was particularly important in the economic context in which McCarthy wrote. Post-war markets were shaped by expanding supermarkets, national retailers, road transport, warehousing and branded consumer goods. A product could not succeed if customers could not find it at an appropriate time and location.

Digital commerce appeared to make Place disappear because a website or app could be accessed from almost anywhere. In reality, Place became more complex. Search engines, app stores, social platforms, digital marketplaces, warehouses, fulfilment centres and last-mile delivery all became parts of distribution.

Instagram Shopping, TikTok Shop and creator affiliate links show how social media transformed Place into a communication environment. A product can be discovered, demonstrated and purchased within or through the same platform.

The social-media feed is therefore not only promotional space. It can also be a point of distribution and sale.

Promotion: One Element, Not the Whole of Marketing

Promotion includes the methods used to inform, persuade and remind audiences. Advertising, personal selling, sales promotion, publicity, public relations, sponsorship and direct communication belong within this category.

A persistent misunderstanding equates marketing with Promotion. Companies sometimes use “marketing” to mean advertisements, social posts or brochures. The 4Ps framework was partly valuable because it made clear that communication was only one part of marketing.

A weak product cannot be repaired permanently through promotion. An inappropriate price or inadequate distribution system may prevent success regardless of advertising creativity.

Historically, Promotion expanded alongside mass media. Newspapers, magazines, radio and television allowed companies to address large audiences. Advertising agencies developed specialised expertise in media buying, copywriting and visual communication.

Social media changed Promotion by making communication interactive, measurable and participatory. Brands publish their own content, users respond publicly and creators integrate products into personal narratives. Paid, owned and earned media overlap.

A social-media campaign may also affect the other Ps. A creator may participate in product development, provide an affiliate discount, sell through a platform and distribute the product digitally. What appears to be Promotion can therefore incorporate Product, Price and Place.

Why McCarthy’s Framework Became Dominant

The history of ideas is not determined only by originality. Some concepts succeed because they solve communication and educational problems exceptionally well.

McCarthy’s framework had several advantages. It was concise. The categories were sufficiently broad to accommodate many decisions. The alliteration made them easy to remember. They fitted a managerial planning process and could be presented visually.

Basic Marketing also became a highly influential textbook. Repeated editions exposed generations of students to the model. Graduates carried the framework into companies, agencies and universities. Once widely adopted, the 4Ps became self-reinforcing: lecturers taught them because textbooks used them, textbooks used them because lecturers and practitioners recognised them, and practitioners recognised them because they had learned them at university.

Silverman’s CHARM research on the development of the marketing-mix concept demonstrates that diffusion through education and professional communities was essential to its institutionalisation (Silverman, 1995).

The framework also arrived at a favourable historical moment. Marketing management was becoming a distinct organisational responsibility. Companies needed tools for planning and coordination. The 4Ps offered a common language across product, pricing, distribution and communication teams.

Philip Kotler: Populariser, System Builder, but Not the Inventor

Philip Kotler is frequently described online as the inventor of the 4Ps. This is historically incorrect.

McCarthy published the four-P classification in 1960. Kotler began writing Marketing Management: Analysis, Planning, and Control in 1963, and the first edition appeared in 1967. Kotler’s book became extraordinarily influential and helped establish marketing management as a systematic discipline. Kotler himself has described beginning the manuscript in 1963 and its publication four years later (Kotler, 2024).

Kotler incorporated the marketing mix into a broader planning framework involving market analysis, segmentation, targeting, positioning, implementation and control. This integration was crucial.

The four Ps alone tell managers which broad instruments can be adjusted. They do not determine which customers should be served or what position the brand should occupy. Kotler’s managerial system connected the mix to strategic analysis.

The familiar sequence became:

analyse opportunities, segment the market, select target groups, determine positioning and configure an appropriate marketing mix.

Kotler also helped internationalise the language of marketing. Numerous editions and translations of Marketing Management brought the framework into classrooms and companies worldwide.

His role was therefore comparable to that of an influential system builder and global educator. He did not create McCarthy’s categories, but he greatly extended their reach and managerial relevance.

The Marketing Mix and the Rise of the Marketing-Management Paradigm

The dominance of the 4Ps reflected a larger shift in marketing thought. Marketing increasingly came to be understood from the perspective of the firm and its manager.

This managerial paradigm focused on decisions the organisation could influence. It offered practical tools for achieving objectives in selected markets.

Ronald Fullerton challenged simplistic histories that presented marketing as progressing through neat production, sales and marketing eras (Fullerton, 1988). Nevertheless, the post-war period did witness growing interest in marketing planning, consumer research, segmentation and managerial control.

The 4Ps were especially compatible with large manufacturing firms selling branded goods. Such firms developed products, established national prices, negotiated distribution and invested in mass-media promotion.

The framework was consequently shaped by its institutional context. This does not invalidate it, but it helps explain its limitations when applied to services, relationships, non-profit organisations or digital ecosystems.

Criticism of the 4Ps

The 4Ps have attracted extensive criticism. One common argument is that they represent the seller’s perspective. Product asks what the company offers, Price what it charges, Place where it distributes and Promotion how it communicates.

Customers may experience the same market differently. They evaluate solutions, total costs, convenience, trust and dialogue.

Another criticism concerns the separation of the four categories. In practice, decisions overlap. Packaging may belong to Product, but also communicates and therefore performs a promotional function. A retail environment belongs to Place but shapes brand perception. Price itself sends a quality signal.

Van Waterschoot and Van den Bulte argued that the 4P classification lacked a sufficiently rigorous basis and contained conceptual overlaps (Van Waterschoot and Van den Bulte, 1992). Their influential reassessment demonstrated that widespread use did not eliminate the need for theoretical scrutiny.

Christian Grönroos criticised the marketing-mix paradigm for encouraging a transactional and internally oriented view of marketing, particularly in services and relationship contexts (Grönroos, 1994). If marketing is reduced to manipulating variables, customers may be treated as passive targets rather than participants in long-term relationships.

The framework can also reinforce departmental isolation. A marketing department may be assigned Promotion while product development, pricing and distribution remain elsewhere, despite all four supposedly forming the marketing mix.

Mark Tadajewski’s critical marketing history offers a broader warning: managerial frameworks should not be presented as neutral, universal truths. They reflect particular institutional interests and assumptions about markets, consumers and organisational control (Tadajewski, 2016).

From 4Ps to 7Ps: Services Change the Marketing Mix

The growth of service industries exposed limitations in a framework developed mainly around goods marketing.

Services are frequently produced and consumed through interactions among customers, employees and systems. Their quality can depend on behaviour, environment and process rather than on a physical object alone.

In 1981, Bernard Booms and Mary Jo Bitner proposed an expanded services marketing mix. They added three elements:

People, Process and Physical Evidence.

People includes employees, customers and others who influence service delivery. Process covers the systems, sequence and procedures through which the service is provided. Physical Evidence includes the tangible signals customers use to evaluate an otherwise intangible offering, such as buildings, uniforms, documents, interfaces or environmental design.

The model first appeared in their contribution to an American Marketing Association conference volume on services marketing (Booms and Bitner, 1981).

The three additions remain highly relevant to digital and social-media services. A platform’s employees and creators affect the user experience. Algorithms and moderation policies are parts of Process. Interface design, reviews, verification marks and profile presentation provide Physical Evidence.

The 7Ps therefore did more than add items to a list. They recognised that marketing may be inseparable from operations and service delivery.

The 4Cs: Reframing the Mix from the Customer’s Perspective

Robert Lauterborn proposed the 4Cs in 1990 as a customer-oriented reinterpretation of the marketing mix.

Product became Customer solution. Price became Customer cost. Place became Convenience. Promotion became Communication.

The 4Cs did not necessarily replace every managerial function contained in the 4Ps. They changed the perspective from which decisions were considered.

A Product focus may encourage a company to begin with what it can manufacture. Customer solution begins with a problem or desired outcome.

Price concentrates on the monetary amount charged. Customer cost includes time, risk, effort and other sacrifices.

Place focuses on distribution controlled by the seller. Convenience asks how easily the customer can access and use the offering.

Promotion can imply one-way persuasion. Communication emphasises dialogue and feedback.

This reinterpretation became especially attractive in relationship and digital marketing, where customers can respond publicly and move between numerous channels.

Nevertheless, the 4Cs are not the original 4Ps and should not be confused with their history. They belong to the continuing debate about how the mix should evolve.

The Marketing Mix in the Age of Social Media

Social media has not abolished the marketing mix. It has changed the meaning and interaction of its elements.

Product in social media

Products are now surrounded by visible user experiences. Reviews, unboxing videos, tutorials, complaints and modifications become part of the market offering’s public identity.

Communities may influence design before and after launch. Software developers release updates in response to user feedback. Consumer brands invite followers to vote on colours, flavours or packaging.

The distinction between product and communication becomes less clear. A shareable package, an interactive app feature or a branded digital filter may be part of the product experience and simultaneously generate promotion.

Creators may also become co-producers. A fashion influencer can collaborate on a collection, attach personal reputation to it and introduce it to an existing community. The creator is not only a promotional intermediary but part of product development and branding.

Price in social media

Social platforms increase price transparency. Consumers compare offers, discuss discounts and publish reactions to price changes.

Affiliate links, creator codes and limited-time offers make price part of social communication. Different creators may distribute different incentives, making pricing measurable at the level of individual partnerships.

Digital business models also complicate monetary price. Social networks offer apparently free access while monetising attention and data through advertising. Customers pay with more than money.

Dynamic pricing, subscriptions, memberships and in-app purchases require marketers to think beyond a single listed amount.

Place in social media

Social media has become a distribution environment. Product discovery, information and transaction may take place within one platform.

A creator video can contain a product link. An Instagram post can carry a shopping tag. TikTok Shop combines entertainment, recommendation and checkout. Livestream commerce turns demonstration and selling into a real-time social event.

Place now includes algorithms, platform interfaces, fulfilment systems and digital marketplaces. The customer’s route to a product may begin in an entertainment feed rather than a store or search engine.

Promotion in social media

Promotion has become participatory. Brands no longer control every public message. Users remix campaigns, create reviews, criticise claims and build alternative meanings.

Influencer marketing embeds commercial communication in personal content. User-generated content allows customers to become distributors of brand imagery. Algorithms determine which promotional messages obtain visibility.

Promotion is therefore no longer a simple transmission from firm to audience. It is a negotiated process involving brands, creators, users and platforms.

Instagram as an Example of the Integrated 4Ps

Instagram illustrates why the four Ps remain useful but difficult to separate.

Consider a beauty brand working with an influencer to launch a cosmetic product. The influencer may participate in selecting the product shade or packaging, contributing to Product. A personalised discount code affects Price. The product can be purchased through a tagged post or connected shop, making Instagram part of Place. The influencer’s Reel, Story and recommendation perform Promotion.

One collaboration activates all four Ps.

The platform also shapes each decision. Its visual norms influence packaging. Its metrics affect promotional content. Its commerce tools structure distribution. Its data may inform pricing and product development.

This demonstrates a major transformation since McCarthy’s period. The marketing manager no longer mixes only internal organisational ingredients. The mix is co-produced with platforms, creators, distributors and consumers.

Borden’s original metaphor may therefore be more useful than a rigid interpretation of four separate boxes. The manager still mixes ingredients, but many of those ingredients are controlled partially by other actors.

The Difference Between the Marketing Mix and Marketing-Mix Modelling

A modern terminological distinction is necessary. The classic marketing mix refers to the managerial configuration of Product, Price, Place and Promotion.

Marketing-mix modelling, commonly abbreviated as MMM, is a statistical approach used to estimate how different marketing activities contribute to sales or other outcomes. It often analyses advertising spending across television, search, social media and other channels.

The two concepts are historically related through the idea of combining marketing inputs, but they are not identical.

A company may use the 4Ps to plan its strategy and use an MMM model to estimate the effects of media expenditure. Contemporary statistical models often concentrate heavily on promotional allocation rather than the complete Product–Price–Place–Promotion system.

Confusing the concepts can lead marketers to treat the marketing mix as only a media budget. Historically, this reverses one of McCarthy’s most important lessons: advertising is only one element of marketing.

Why the 4Ps Have Survived

Many alternative frameworks have been proposed, yet none has displaced the 4Ps completely.

Their persistence is partly explained by simplicity. The model provides an accessible starting point for discussing a market offer. It can be applied to a consumer product, service, charity or digital platform with suitable interpretation.

The categories also describe real managerial responsibilities. Organisations still decide what to offer, what exchange to request, how to provide access and how to communicate.

The model is flexible enough to absorb new practices. E-commerce remains a Place decision. Influencer communication belongs partly to Promotion. Subscription pricing remains Price. Digital services remain Products or offerings.

The 4Ps survive not because markets have remained unchanged, but because their categories are broad.

Their weakness is the reverse side of this strength. Categories broad enough to include almost everything may explain relatively little without additional analysis.

The framework therefore works best as a checklist or organising language, not as a complete theory of customer behaviour, competition or society.

Who Really Invented the 4Ps?

The question can now be answered accurately.

James W. Culliton did not invent the 4Ps. In 1948, he described the marketing executive as a mixer of ingredients.

Neil H. Borden did not formulate Product, Price, Place and Promotion as four categories. He transformed Culliton’s metaphor into the concept of the marketing mix, used the term from the late 1940s onward and published an influential explanation in 1964.

E. Jerome McCarthy created the 4P classification and published it in Basic Marketing: A Managerial Approach in 1960.

Philip Kotler did not invent the concept or the classification. His Marketing Management, first published in 1967, incorporated the marketing mix into a comprehensive managerial planning system and contributed enormously to its worldwide dissemination.

Booms and Bitner extended the model to seven Ps for services in 1981.

Lauterborn reframed it through four customer-oriented Cs in 1990.

The history is collaborative and cumulative. Reducing it to one inventor hides the intellectual process through which marketing became a management discipline.

The Broader Historical Contribution of the Marketing Mix

The marketing mix changed marketing thought in at least three important ways.

First, it made coordination central. Product, pricing, distribution and communication should support the same market position.

Second, it established marketing as a managerial responsibility extending beyond advertising and selling. Marketing influenced the offer itself and the means by which customers obtained it.

Third, it offered a teachable bridge between analysis and action. Segmentation and customer research identify opportunities; the marketing mix translates strategic choices into operational decisions.

Its limitations are equally instructive. Markets cannot be controlled like laboratory experiments. Customers are not passive recipients. Institutions, cultures, competitors and technologies shape outcomes. Relationships and services often require categories beyond the original four.

Borden’s emphasis on environmental forces and managerial judgement reminds us that the mix was never meant to function mechanically.

Conclusion

The four Ps are among the most recognisable ideas in the history of marketing, but their familiar simplicity conceals a layered intellectual development.

Their story began with James Culliton’s 1948 description of the marketing executive as a mixer of ingredients. Neil H. Borden adopted that metaphor and developed the marketing-mix concept as an extensive set of managerial variables shaped by consumers, competitors, intermediaries and regulation.

E. Jerome McCarthy achieved the decisive simplification. In 1960, he grouped the manager’s controllable marketing variables under Product, Price, Place and Promotion. The alliterative structure made a complicated field easier to teach and apply.

Philip Kotler subsequently embedded the 4Ps within a broader system of analysis, segmentation, targeting, positioning, planning and control. His influential textbooks brought the model to international generations of students and managers. Kotler should therefore be recognised as one of its most important popularisers and system builders, but not as its original inventor.

The framework reflected the post-war rise of professional marketing management and branded mass markets. Its perspective was primarily that of the firm. This made it practical, but also generated criticism. Services scholars added People, Process and Physical Evidence. Customer-oriented thinkers reframed the Ps as solution, cost, convenience and communication. Relationship and critical marketing scholars questioned the image of customers as targets of controllable managerial instruments.

Social media has made these limitations more visible while confirming the continuing relevance of the underlying decisions. A brand on Instagram must still decide what it offers, how value is exchanged, how the offer can be accessed and how it will communicate. Yet those decisions are increasingly interconnected and shared with creators, platforms and communities.

Product can become content. Promotion can become distribution. Place can become an algorithmic feed. Price can become a creator code, subscription or exchange of personal data.

For this reason, the four Ps should neither be discarded as obsolete nor treated as a complete definition of marketing. They remain a durable map of managerial decisions, but the territory has changed.

The most historically appropriate interpretation returns to Borden’s original metaphor. Marketing is not the mechanical activation of four independent levers. It is the continuous mixing of interdependent ingredients under uncertain market conditions.

McCarthy gave those ingredients four memorable containers. Kotler taught managers how to connect them to strategy. Digital platforms have now mixed them together again.

References

Aaker, D.A. (1991) Managing Brand Equity: Capitalizing on the Value of a Brand Name. New York: Free Press.

Bartels, R. (1988) The History of Marketing Thought. 3rd edn. Columbus, OH: Publishing Horizons.

Berghoff, H. (ed.) (2007) Marketinggeschichte: Die Genese einer modernen Sozialtechnik. Frankfurt am Main and New York: Campus.

Booms, B.H. and Bitner, M.J. (1981) ‘Marketing strategies and organization structures for service firms’, in Donnelly, J.H. and George, W.R. (eds.) Marketing of Services. Chicago: American Marketing Association, pp. 47–51.

Borden, N.H. (1964) ‘The concept of the marketing mix’, Journal of Advertising Research, 4(2), pp. 2–7.

Culliton, J.W. (1948) The Management of Marketing Costs. Boston, MA: Division of Research, Graduate School of Business Administration, Harvard University.

Fullerton, R.A. (1988) ‘How modern is modern marketing? Marketing’s evolution and the myth of the “production era”’, Journal of Marketing, 52(1), pp. 108–125.

Grönroos, C. (1994) ‘From marketing mix to relationship marketing: Towards a paradigm shift in marketing’, Management Decision, 32(2), pp. 4–20.

Holt, D.B. (2004) How Brands Become Icons: The Principles of Cultural Branding. Boston, MA: Harvard Business School Press.

Jones, D.G.B. and Tadajewski, M. (eds.) (2016) The Routledge Companion to Marketing History. Abingdon and New York: Routledge.

Kapferer, J.-N. (2012) The New Strategic Brand Management: Advanced Insights and Strategic Thinking. 5th edn. London: Kogan Page.

Keller, K.L. (1993) ‘Conceptualizing, measuring, and managing customer-based brand equity’, Journal of Marketing, 57(1), pp. 1–22.

Kotler, P. (1967) Marketing Management: Analysis, Planning, and Control. Englewood Cliffs, NJ: Prentice-Hall.

Kotler, P. and Keller, K.L. (2016) Marketing Management. 15th edn. Harlow: Pearson.

Lauterborn, R.F. (1990) ‘New marketing litany: 4Ps passé; C-words take over’, Advertising Age, 61(41), p. 26.

Lazer, W. and Kelley, E.J. (1962) Managerial Marketing: Perspectives and Viewpoints. Homewood, IL: Richard D. Irwin.

McCarthy, E.J. (1960) Basic Marketing: A Managerial Approach. Homewood, IL: Richard D. Irwin.

Petty, R.D. (2016) ‘A history of brand identity protection and brand marketing’, in Jones, D.G.B. and Tadajewski, M. (eds.) The Routledge Companion to Marketing History. Abingdon and New York: Routledge.

Shaw, E.H. (2016) ‘Ancient and medieval marketing’, in Jones, D.G.B. and Tadajewski, M. (eds.) The Routledge Companion to Marketing History. Abingdon and New York: Routledge.

Silverman, S.N. (1995) ‘An historical review and modern assessment of the marketing mix concept’, Proceedings of the Conference on Historical Analysis and Research in Marketing, 7, pp. 25–35.

Tadajewski, M. (2016) ‘Critical marketing history’, in Jones, D.G.B. and Tadajewski, M. (eds.) The Routledge Companion to Marketing History. Abingdon and New York: Routledge.

Tamilia, R.D. (2016) ‘A history of channels of distribution’, in Jones, D.G.B. and Tadajewski, M. (eds.) The Routledge Companion to Marketing History. Abingdon and New York: Routledge.

Van Waterschoot, W. and Van den Bulte, C. (1992) ‘The 4P classification of the marketing mix revisited’, Journal of Marketing, 56(4), pp. 83–93.

Wernick, A. (1991) Promotional Culture: Advertising, Ideology and Symbolic Expression. London: Sage.