A handful of paper slips could determine whether an advertising campaign deserved a national budget. They carried orders, requests for samples and identifying marks that connected a response to a particular advertisement. For Claude C. Hopkins, these modest documents offered something that an enthusiastic meeting could not: evidence of what prospective customers had actually done. The next question was financial. How much had each acquired customer cost?
Published in 1923, Scientific Advertising turned this approach into a compact programme for advertising practice. Hopkins connected copywriting, selling, customer knowledge, distribution and test campaigns with commercial accountability. His historical contribution was not the invention of every technique he described. It was the forceful argument that advertising decisions should be open to examination and that recorded experience should guide subsequent work (Hopkins, 1923, ch. 1).
The book belongs to the prehistory of performance marketing and conversion optimisation. That does not make Hopkins the sole inventor of measurable advertising or the originator of modern statistical A/B testing. Reading him closely reveals something more useful: a practical economics of customer acquisition, accompanied by ambitious claims about scientific certainty that deserve scrutiny of their own.
What Scientific Advertising promised in 1923
Hopkins began with an unusually confident proposition. Advertising, when directed by capable practitioners, had reached the status of a science. Causes and effects had been investigated, procedures compared and basic principles established. An activity previously treated as a gamble could therefore become a relatively safe commercial undertaking (Hopkins, 1923, ch. 1).
This opening was both a statement of method and a persuasive case for professional advertising. Hopkins wanted business owners to believe that advertising expenditure did not have to be an obscure wager. Agencies could draw on documented experience instead of restarting every assignment with personal preferences and unsupported opinions. The central resource was accumulated knowledge, obtained through comparison and made available for reuse.
His method and his rhetoric nevertheless need to be separated. Repeated comparisons can improve decisions. They do not establish that advertising follows universal natural laws. Hopkins moved between a valuable demand for empirical investigation and a much stronger promise of predictable human behaviour. That tension is essential to understanding the book. Scientific Advertising is a historical argument about how business should know what works, rather than an inventory of rules whose authority can simply be assumed.
The advertising agency as an institutional memory
One of the book’s most revealing passages concerns organisations rather than headlines. Hopkins described large agencies as repositories of advertising experience. Employees might leave, but their records, ideas and discoveries should remain. Subsequent staff could then work with knowledge produced by campaigns they had never personally handled (Hopkins, 1923, ch. 1).
Scientific advertising therefore depended partly on institutional learning. An agency was more than a supplier of copy and purchased space. It could compare results, preserve findings and transfer lessons between assignments. Under those conditions, an unsuccessful trial might still provide useful information, provided that its circumstances and outcome were properly recorded.
This can be interpreted as an early organisational logic of experimentation. The expression is a present-day interpretation, not Hopkins’ own terminology. The distinction matters because an archive of attractive advertisements is not the same thing as an archive of useful evidence. Costs, response mechanisms, market conditions and unsuccessful alternatives determine what can be learned. Hopkins envisaged a commercial organisation whose experience survived individual careers and became a resource for future decisions.
Advertising as multiplied personal selling
In his second chapter, Hopkins presented the perspective that shaped the whole book: advertising was selling. An advertisement should therefore face standards comparable to those applied to a salesperson. Knowledge of the product, understanding of the customer, clarity of explanation and an economically satisfactory result mattered more than literary distinction (Hopkins, 1923, ch. 2).
The crucial difference was scale. A salesperson addressed one prospect at a time, while an advertisement could repeat a proposition before thousands. That multiplication increased opportunity, but it also increased the consequences of a mistake. An ineffective argument became more expensive when distributed across a large campaign.
This explains his suspicion of copy that chiefly displayed its author’s cleverness. Hopkins was not recommending incoherent or impoverished writing. He wanted the writer’s skill to serve the buyer’s understanding. Originality had to justify its commercial role. This remained a narrower position than a comprehensive account of brand communication, which might also consider cultural meanings, reputation and effects that develop over long periods. His choice of the salesperson as the governing analogy shaped both the strengths and the limitations of his approach.
Mail-order advertising as a practical testing environment
Hopkins explicitly credited mail-order advertising with establishing many of his basic principles. In that environment, a business could connect an advertisement relatively directly with incoming orders. Expenditure and responses could be assigned to particular placements. Even small improvements mattered when repeated across large circulations and successive campaigns (Hopkins, 1923, chs 1 and 4).
This changed how creative decisions were evaluated. Advertisement size, illustration, headline and offer could be compared rather than judged entirely by taste. A larger advertisement needed to produce enough additional business to justify the extra space. A smaller one was not necessarily economical if it failed to explain the proposition adequately.
Mail order did not, however, reproduce every advertising situation. An immediately orderable product offers different measurement opportunities from an article purchased much later through a retailer. Hopkins extended the discipline of direct response to other settings, but the extension required additional assumptions. The commercial visibility of one channel could not by itself provide a complete explanation of all advertising effects. His achievement was to make a demanding standard of accountability portable, even where the original measurement environment could not be reproduced exactly.
Keyed advertisements and the attribution of responses
The practical foundation frequently involved identifying marks. A coupon, order form or other response device could indicate which advertisement had generated an enquiry. Giving different versions different identifiers allowed the advertiser to connect an incoming response with its source (Hopkins, 1923, ch. 1).
The coupon was consequently more than a discount instrument. It established an observable connection between published advertising and a subsequent action. Readers might request a sample, ask for information or redeem an offer at a shop. A public message now had a response mechanism that could be counted and compared.
There is an obvious resemblance to campaign parameters and dedicated destination addresses. The resemblance concerns attribution first, however, rather than a complete proof of causation. Even a perfectly identified order does not show whether the customer would have purchased without seeing the advertisement. Hopkins made commercial responses more visible. That was an important contribution, but it did not automatically establish the amount of additional demand caused exclusively by advertising. Keeping those questions separate prevents a useful historical comparison from becoming an exaggerated claim about modern measurement.
Why more replies did not necessarily mean better advertising
Hopkins warned against making decisions from response totals alone. One advertisement could generate many replies with little commercial value, while another produced fewer responses from more valuable buyers. Final judgements should therefore depend on the cost of obtaining customers or the relationship between advertising cost and sales (Hopkins, 1923, ch. 1).
This is more sophisticated than a historical version of counting clicks. Hopkins understood that an easily stimulated intermediate action might diverge from a satisfactory business outcome. A free offer could attract numerous people who would never become paying customers. Another proposition might generate fewer enquiries but ultimately produce a stronger commercial result.
The modern distinction between lead volume and lead quality provides a helpful analogy. It does not mean that Hopkins had already developed today’s lead-management terminology. His underlying question was more basic: which responses were worth paying to obtain? This placed the economic quality of demand at the centre of campaign evaluation. It also challenged the temptation to celebrate a large, immediately visible number simply because it was easy to measure and impressive to report.
Headlines as a means of selecting the right audience
Hopkins treated a headline as more than an ornamental introduction. Its function was to call out the people for whom the proposition mattered. An advertisement in a mass publication did not need to interest everybody. It needed to identify suitable prospects within the wider readership (Hopkins, 1923, ch. 5).
His discussion of soap illustrates the point. Cleanliness, composition, the ability to float and benefits associated with appearance represented different routes into the same broad product category. Testing could reveal which benefit attracted the most useful interest. The exercise therefore extended beyond changing a word or rearranging a sentence. It could investigate the central reason for considering the product.
Hopkins also recognised that an appeal producing fewer responses might still reach a profitable group. A weaker overall result did not necessarily make a proposition worthless. This matters historically because it complicates the idea that testing always discovers one universal message. Multiple commercially viable approaches could coexist. His argument touched questions of audience differentiation without amounting to a fully developed version of later segmentation theory.
Specific information and the persuasive appearance of precision
A separate chapter defended specificity. Broad assertions of superiority were easy for readers to dismiss as routine self-praise. A definite statement appeared different because it committed the advertiser to a particular claim rather than a general impression (Hopkins, 1923, ch. 7).
This recommendation depended on investigation. To communicate a meaningful detail, the writer first had to discover and understand it. Manufacturing processes, product characteristics and patterns of use became material for advertising. The persuasive proposition should emerge from knowledge of the article rather than being attached to it as an unrelated flourish.
Precision nevertheless needs its own examination. A specific number is not automatically true, and numerical detail can mislead when its basis is missing. Hopkins’ argument therefore contains two distinguishable elements: the value of substantiated information and the rhetorical force of apparent exactness. The first improves the buyer’s knowledge; the second can create unwarranted confidence. His scientific vocabulary makes that distinction especially important. An advertisement does not become reliable merely because it replaces an adjective with a figure.
Puffed Wheat and Puffed Rice: manufacturing becomes a story
Hopkins described the promotion of Puffed Wheat and Puffed Rice through a striking account of their production. The grain was presented as food shot from guns and expanded to several times its ordinary size. Curiosity transformed a familiar category into something unusual and worth investigating (Hopkins, 1923, ch. 6).
The example demonstrates how misleading it would be to reduce his approach to dry calculation. A technical process became a dramatic mental picture. Information and imagination worked together: the process gave the proposition substance, while the language made that process memorable. Scientific advertising did not mean abandoning an arresting story. It meant wanting the story’s commercial performance to be examined.
The source also calls for caution. Hopkins himself attributed commercial success to this approach. At that point, his book does not provide independently analysable sales records that isolate the contribution of the wording from distribution, product changes or other influences. The passage reliably documents what he considered persuasive and how he explained his work. It does not, by itself, prove an exact causal contribution to revenue. That distinction protects the historical example from becoming an unqualified success legend.
Research before writing the advertisement
The chapter entitled Information challenges the idea that Hopkins’ method was simply a collection of headline tricks. He demanded substantial preparation: reading specialist literature, investigating competitors, consulting dealers and finding out how people actually used or evaluated an offer. A useful campaign idea might emerge from one relevant fact located among many unhelpful pages (Hopkins, 1923, ch. 11).
His account of canned pork and beans illustrates a change in the definition of competition. According to Hopkins, investigation showed that the major alternative was not another canned brand. Many households prepared beans themselves. Advertising therefore had to explain why the prepared category deserved consideration before it could effectively argue for a particular manufacturer.
The strategic point is important. Businesses can misunderstand their market by examining only competitors that resemble themselves. The customer’s existing behaviour may be the more significant alternative. Hopkins supplied numerical claims for this case without the research documentation needed for independent modern verification. The appropriate reading is therefore a contemporary practitioner’s report, rather than an audited market estimate. Its historical value lies in showing how customer investigation could overturn the problem that management initially thought it needed to solve.
Complete arguments rather than a universal rule about length
Hopkins is often associated with long advertising copy. His own argument was more conditional. An advertising story should be sufficiently complete for the product and decision involved. A simple article might need very little explanation. An important purchase could warrant considerably more information (Hopkins, 1923, ch. 8).
He described a motor-car buyer who selected a Marmon after encountering more substantial information than competing manufacturers supplied. The episode challenged the assumption that affluent or sophisticated customers necessarily preferred short advertisements. Someone considering an expensive purchase might willingly read extensive material if it answered relevant questions.
The resulting principle is not that longer copy always sells better. It is that the advertiser should avoid leaving decisive questions unanswered. Applied as a historical analogy to a contemporary product page, the idea favours useful completeness rather than maximum length. Brevity can improve understanding, but it can also remove the information that makes a purchase possible. Nothing in this argument establishes a universally successful word count. The relationship between the decision and the information required matters more than a fixed preference for either short or long text.
Samples and the effort required to respond
Hopkins paid close attention to the practical work involved in responding. A person might need to cut out a coupon, write a letter, find a stamp or visit a shop. These differences could affect results even when the product and central proposition remained unchanged (Hopkins, 1923, ch. 13).
He reported one case in which postal sample enquiries cost an average of 70 cents each, while the same advertisements generated enquiries for 18 to 22 cents when coupons were presented locally. In another case offering a choice between writing, telephoning and visiting a shop, he stated that 70 per cent of enquiries arrived by telephone. These are Hopkins’ reported campaign figures, not general historical benchmarks or universally reproducible findings.
Their significance lies in what he chose to notice. The route from interest to action formed part of advertising performance. Contemporary conversion optimisation examines comparable obstacles in registration, forms and checkout. The historical continuity concerns the cost of action to the prospective customer. Being persuaded is not sufficient if completing the next step demands too much effort. For Hopkins, practical convenience belonged within the analysis of selling rather than outside it.
Distribution and the problem of unavailable products
Another chapter addressed distribution directly. Hopkins considered extensive advertising commercially questionable if the people persuaded by it could not subsequently obtain the product. The reach of communication needed to correspond with the reach of supply (Hopkins, 1923, ch. 14).
He described how forthcoming coupon promotions could encourage dealers to carry stock. Sending a retailer a proof of the advertisement and information about its planned circulation made anticipated customer demand tangible. The promotion connected consumer acquisition with the problem of securing retail availability. Hopkins named Palmolive and puffed-grain products among the examples of this approach.
This broadens his place in marketing history beyond better copywriting. The advertisement’s effectiveness depended partly on the business’s ability to serve the demand it created. In modern terms, one might discuss coordination between acquisition and operational capacity, but the historical relationship is straightforward: the advertisement, the sample and the selling point belonged to a single commercial arrangement. A response mechanism was useful only if the interested person could continue towards a purchase. Measurement of the advertisement alone could conceal failure elsewhere in that arrangement.
Test markets, repeat purchases and the financing of growth
In Test Campaigns, Hopkins contrasted a limited market trial with an immediate national commitment. A business could begin in a few towns and investigate acquisition costs, actual buying and subsequent repeat purchases before expanding. The purpose was to reduce the scale of an initial mistake and obtain information before committing larger resources (Hopkins, 1923, ch. 15).
Time was part of the calculation. Advertising expenditure might return through earnings only after several months. A campaign could therefore look profitable while still creating a financing requirement. The business needed enough money to support the interval between spending and recovery.
This anticipates questions now discussed through payback periods and customer value. Hopkins did not supply a complete contemporary customer-lifetime-value model. He did, however, connect repeat buying, earnings and the recovery of acquisition expenditure. That connection is central to the book’s relevance for performance-marketing history. The initial response was not always the end of the analysis. A customer could be expensive to acquire but valuable over time, while an apparently attractive first transaction might fail to develop into a viable commercial relationship.
Why Hopkins did not simply invent the modern A/B test
Comparing alternatives is not identical to conducting a fully controlled experiment. Advertisements may run in different publications, towns or weeks. Their audiences, competitive environments, dealer coverage and seasonal conditions may consequently differ as well. A superior result cannot automatically be assigned to the changed headline alone.
Scientific Advertising described repeated testing and financial comparisons, but it did not present an integrated protocol of random allocation, predetermined statistical decision criteria and reported uncertainty intervals. Calling Hopkins the inventor of modern randomised A/B testing would therefore extend the evidence beyond what his book establishes (Hopkins, 1923, chs 1 and 15).
His confidence about scaling also deserves examination. He suggested that learning the cost of acquiring a thousand customers could reveal almost exactly what a million would cost. Expansion can instead reach less suitable audiences, require more expensive media or encounter limited demand. This is a methodological criticism of his generalisation, not an argument against running small trials. Commercial experience and statistically supported causal inference are different forms of knowledge. Appreciating Hopkins’ contribution does not require treating them as interchangeable.
Earlier and international comparisons
Walter Dill Scott published The Theory of Advertising in 1903, connecting psychological principles with successful advertising two decades before Hopkins’ book. Its existence alone challenges the suggestion that systematic interest in advertising effects began in 1923 (Scott, 1903).
William A. Shryer offers a closer predecessor. Tadajewski (2020) examines his early combination of advertising analysis, habitual behaviour and psychological explanation. The study widens the history beyond Hopkins: emerging advertising science involved multiple contributors and addressed forms of behaviour that could not be reduced to deliberate rational calculation.
For Britain, Schwarzkopf (2007) shows how Lever and J. Walter Thompson used consumer investigation between 1927 and 1935 in developing Lux. This was a neighbouring path rather than an earlier version of Hopkins’ book. Market knowledge involved studying use and consumer needs as well as counting responses.
The international and long historical scope of The Routledge Companion to Marketing History, edited by Jones and Tadajewski (2016), likewise cautions against concentrating the field around a single American founder. These comparisons do not establish a worldwide ranking of the first advertising experiments. They demonstrate why such a priority claim would require a much broader evidential base. Hopkins belongs within a developing field of practice and inquiry, rather than standing outside history as its solitary point of origin.
Brand personality and what immediate response cannot capture
Despite his sales orientation, Hopkins also discussed familiarity and trust. In Individuality, he recommended an appropriate and consistent tone. Continually changing the advertiser’s personality made it harder for people to become acquainted with the business. A recognisable manner should persist across successive advertisements (Hopkins, 1923, ch. 17).
This sits uneasily with the simplified image of an author interested exclusively in the immediate return from an isolated advertisement. Hopkins understood that communication could establish expectations over time. Nevertheless, he did not offer a complete method for separating the contribution of that accumulated familiarity from short-term responses.
The resulting historical question remains useful: what disappears when only easily counted actions are accepted as effects? Advertising may prepare later demand without generating an identifiable reply. Conversely, a promotion may obtain a fast response while weakening trust. Hopkins’ demand for accountability remains valuable, but the measurement horizon helps determine which version of success becomes visible. A longer relationship and an immediate order need not tell the same story. Responsible evaluation requires acknowledging that difference rather than assuming that one convenient number captures everything advertising does.
Scientific language, persuasion and responsibility
The language of science performed a persuasive role inside Scientific Advertising itself. References to laws, proof and exactness gave authority to the proposed method. Hopkins offered both practical procedures and the hope that commercial uncertainty could be brought under control (Hopkins, 1923, chs 1 and 15).
A critical reading must distinguish evidence of a commercial response from evidence that a product claim is true. Increased orders do not validate a medical or technical assertion. Equally, the book’s psychological assumptions and social stereotypes remain historical positions; their alleged sales effectiveness cannot supply an adequate justification for them.
Hopkins did call for questionable claims to be investigated, including through specialist expertise where appropriate. That requirement belongs within his programme and should not be overlooked (Hopkins, 1923, ch. 11). Consistent scientific reasoning nevertheless demands more than an impressive figure or a successful appeal. It requires an intelligible basis for the claim, recognition of uncertainty and willingness to abandon an attractive proposition when supporting evidence is absent. His aspiration to reliable knowledge is most useful when applied to his own confident conclusions as well.
What Scientific Advertising contributed to marketing history
Scientific Advertising made a particular working discipline accessible in 1923: formulate a proposition, provide a recognisable route to response, compare expenditure with commercial outcomes and use the findings to guide the next decision. Its connection with later direct marketing and performance marketing lies in this structure of reasoning. It should not be mistaken for an uninterrupted technical genealogy from the printed coupon to every component of a digital advertising platform.
Hopkins remains especially useful when he challenges convenient substitutes for business results. An attractive advertisement is not automatically a good investment. A large number of enquiries does not establish profitable acquisition. An inexpensive first contact does not prove a valuable customer relationship. And a promising local trial does not guarantee unlimited expansion at the same cost.
His historical position is therefore best understood through both achievement and limitation. He organised existing practices into a forceful programme of commercial examination. Taking that programme seriously means testing its rules rather than merely repeating them. The lasting challenge of Scientific Advertising is that an advertising decision should be able to withstand something more demanding than the confidence of the person who proposed it.
References
Hopkins, C.C. (1923) Scientific Advertising. Chicago: Lord & Thomas. Digitised text available at: https://www.scientificadvertising.com/ScientificAdvertising.pdf (Accessed: 21 October 2026). References use chapter numbers because pagination differs between digital reproductions.
Jones, D.G.B. and Tadajewski, M. (eds) (2016) The Routledge Companion to Marketing History. Abingdon: Routledge. Publisher’s description and contents available at: https://www.routledge.com/The-Routledge-Companion-to-Marketing-History-1st-Edition/Jones-Tadajewski/p/book/9781315882857 (Accessed: 21 October 2026).
Schwarzkopf, S. (2007) ‘Consumer voice and brand loyalty: Lever, J. Walter Thompson and the rise of market research in Britain, 1918-1939’, Proceedings of the Conference on Historical Analysis and Research in Marketing, pp. 180-192. Available at: https://ojs.library.carleton.ca/index.php/pcharm/article/view/1583/1427 (Accessed: 21 October 2026).
Scott, W.D. (1903) The Theory of Advertising: A Simple Exposition of the Principles of Psychology in Their Relation to Successful Advertising. Boston: Small, Maynard & Company. Bibliographic record and digitised copy available at: https://books.google.cg/books?id=m9dHAAAAIAAJ (Accessed: 21 October 2026).
Tadajewski, M. (2020) ‘William A. Shryer, scientific advertising, habits and motivation research’, Journal of Historical Research in Marketing, 12(2), pp. 197-218. https://doi.org/10.1108/JHRM-10-2019-0037.
