The checkout counter in Troy carried an agreement that extended far beyond one Ohio supermarket. On 26 June 1974, a package of Wrigley’s Juicy Fruit gum was identified through its Universal Product Code at a Marsh store. Ohio’s historical marker commemorates the event as the first retail scan of a product marked with the UPC. The installation combined a Spectra Physics Model A scanner with an NCR 726 computerised cash register (Ohio History Connection, n.d.). That combination matters: the barcode became useful because a shared identification system could connect the work of different businesses.
George Laurer was central to the development of the rectangular symbol selected for that system. Leading IBM’s engineering response, he helped turn demanding requirements into a practical design. His achievement deserves recognition without making him the sole inventor of every barcode or the author of the entire retail information infrastructure (IBM, n.d.a). The history of the UPC is consequently more revealing than a familiar story about stripes and chewing gum. It explains how engineering, commercial agreement and everyday operating work made a new kind of retail transaction possible.
Separating Invention, Agreement and Application
Three dates belong to different stages of this history. Woodland and Silver’s barcode patent application dates from 1949, with the patent granted in 1952. GS1’s historical timeline records the US grocery industry’s agreement on 3 April 1973 and the first UPC supermarket transaction on 26 June 1974 (Eastman, n.d.; GS1, 2025). These are related events, but treating them as interchangeable creates an inaccurate account of what changed and who contributed.
The earlier invention established an approach to machine-readable identification. The later industry decision established a common basis on which companies could act. The transaction in Troy demonstrated a retail application of that agreed system. RCA had already tested its circular-code system at a Kroger store near Cincinnati in July 1972 (Gerena, 2014). None of those achievements automatically produced the others. A workable idea can remain commercially impractical; a standard can exist before enough businesses implement it; an installation can operate successfully while widespread adoption remains uncertain.
For marketing history, the distinctions expose work that disappears from a simple invention narrative. Packaging has to be prepared, responsibilities allocated and investments justified. The same printed mark must remain useful as a product passes between companies with different systems and priorities. What happened in 1974 was therefore neither the beginning of all optical coding nor the completion of retail automation. It was a particularly visible crossing point between technical development and a commercial practice that could be repeated. Laurer’s contribution becomes clearer when placed at that crossing rather than stretched across the entire preceding and subsequent history.
The Business Problem Was Larger Than a Slow Queue
A supermarket checkout brings together decisions made elsewhere. Manufacturers have already designed their packages. Printers have translated those designs into physical marks. Retailers have selected the assortment and established prices. Equipment suppliers have built devices according to particular assumptions. Improving the checkout means making those decisions work together in a setting where errors and delays immediately affect customers.
IBM describes the initiative around 1970 as a joint call from grocery retailers and packaged-goods manufacturers for a standard machine-readable symbol (IBM, n.d.b). The demand was significant because a retailer-specific solution could transfer difficulties to suppliers. If every chain required its own packaging code, manufacturers would face additional variations to prepare and distribute. Agreement promised to reduce that repeated work, but it also required companies to accept rules that no individual participant could determine entirely alone.
The commercial logic can be understood without assuming that cooperation was easy. A shared convention creates costs before it delivers repeated benefits. Existing procedures need adjustment, staff need instructions and product information needs agreement. Over time, however, the same convention can be used for many transactions and trading relationships. A manufacturer can prepare one compatible marking for numerous customers, while a retailer can process goods from numerous suppliers through a common method. The UPC project addressed this relationship between individual costs and collective usefulness. The symbol was the visible result of a deeper decision about which differences between businesses should remain and which should be removed from routine exchange.
Making an Industry Capable of Choosing
Charles Gerena’s historical account places the first meeting of the ad hoc committee in August 1970. Senior industry representatives took part, with McKinsey supporting the work. A separate symbol-selection group chaired by Alan Haberman evaluated the visual alternatives. Commercial authority and technical assessment were consequently linked within an industry process rather than left to one equipment supplier (Gerena, 2014).
The arrangement addressed an organisational problem common to shared innovations. Engineers can establish that a system works without being authorised to change their company’s packaging or investment policy. Executives can endorse a general ambition without committing to its particular costs. A committee containing people able to secure decisions within their own organisations has a different capacity from a discussion group whose participants must refer every difficult choice elsewhere.
This is an important part of what makes the UPC a marketing achievement. The industry was organising a condition under which transactions could occur more consistently. Cooperation did not remove competition between brands or stores. Instead, it established a common layer beneath that competition. Companies could continue to distinguish their products and services while relying on an agreed method of identifying what was being sold.
Seen this way, the committee was not an administrative footnote to an engineering breakthrough. Its work helped define the problem that engineering had to solve and the circumstances under which a solution would be accepted. A technically impressive device without that acceptance could have remained one more isolated experiment. The agreement gave participants a reason to develop compatible practices beyond the boundaries of the first demonstration.
George Laurer’s Practical Engineering Contribution
Laurer graduated in electrical engineering from the University of Maryland in 1951 and joined IBM that year. By 1969 he was a senior engineer at the company’s Research Triangle Park operation in North Carolina. IBM’s retrospective credits him with challenging the preference for a circular symbol because printing imperfections could compromise its performance. He advanced a rectangular alternative and worked with a team on its implementation (IBM, n.d.a).
This was a practical rather than merely visual decision. A symbol that looks clear in a drawing may behave differently when reproduced on actual packaging. Production introduces variation in materials, surfaces and print quality. A design intended for widespread use has to tolerate the world in which it will be made. The printer and packaging manufacturer therefore belong inside the history of the technology, rather than appearing only after the supposedly important work is finished.
Laurer’s task also involved balancing requirements that could pull in different directions. A large symbol might be easier to recognise but occupy more package space. A dense symbol might carry information compactly while creating difficulties for reading equipment. A successful demonstration could show technical possibility without establishing reliable everyday use. The engineering achievement lay in finding a workable relationship between these requirements.
That interpretation avoids attributing progress to individual insight alone. Laurer’s judgement mattered, but its consequences depended on a team, production capabilities and the industry’s willingness to evaluate and adopt the result. The useful historical question is not simply who drew the winning pattern. It is how that pattern became reproducible and dependable enough to support transactions across many businesses. His contribution was particularly important at the point where design had to meet that demanding commercial environment.
A Rectangular Symbol Did Not Mean a Single Reading Direction
The shape of a barcode can encourage a misleading assumption: a rectangular pattern appears to require exact alignment. Eastman’s technical history instead explains the relationship between UPC construction and scanning patterns that can read different portions of the symbol. Recognition depends on the combined operation of the printed code, optical system and decoding process (Eastman, n.d.). The practical capability belongs to the arrangement as a whole.
For a retailer, that relationship matters because packages do not arrive at the scanner in a perfectly controlled position. The cashier is handling a succession of different objects while maintaining the flow of the transaction. A system that works only after careful positioning may offer less practical benefit than its laboratory performance suggests. Reliability is therefore part of speed. Repeated attempts or corrections can consume the time apparently saved by a fast individual reading.
The point also explains why the barcode cannot be assessed as an independent graphic. Its performance depends on what reads it and on the work into which reading is inserted. A successful symbol must be compatible with ordinary handling as well as ordinary printing. It connects the design of a package to the conduct of the checkout.
This is a distinctive aspect of packaging in marketing history. Packaging must attract and inform a shopper, but it also participates in commercial operations. The UPC added a machine-readable function to that already complex object. Its usefulness cannot be measured only by how it appears on a shelf. It also has to survive the transition from a displayed offer to a recorded sale.
What the 1973 Decision Could and Could Not Accomplish
The selection of the UPC symbol in April 1973 created a shared reference point. It did not instantly equip stores or mark every package. Retailers still needed suitable equipment and operating preparation; manufacturers still needed to implement the marking. GS1 also records the Uniform Code Council’s administrative role in 1974 (GS1, 2025). Management of the system was a continuing responsibility, separate from its initial design.
This distinction helps explain why standards require institutions. A specification can describe how information is represented. It cannot, by itself, guarantee that companies will assign information consistently or maintain it over time. New products have to enter the system. Changes have to be handled without creating ambiguity. Participating businesses need ways of knowing which conventions apply beyond their own internal operations.
Adoption consequently involved more than purchasing machinery. It required companies to make a shared convention part of normal work. The benefits depended on repeated use: one compatible package is useful, but a store gains more when a substantial part of its assortment can be processed in the same way. Manufacturers in turn gain more when additional customers use the system.
The historical significance of Troy should be understood within that process. A first installation can demonstrate that the system operates in a particular setting. It cannot establish how quickly it will spread or how profitable it will be elsewhere. The 1974 event provided a memorable example of practical application. The transformation of retail depended on the much less photogenic work of extending and maintaining that application across products, stores and trading relationships.
Troy Was a Prepared Commercial Demonstration
The association with IBM can obscure an important fact about the first UPC transaction. The installation commemorated in Ohio used equipment from Spectra Physics and NCR (Ohio History Connection, n.d.). Laurer’s role in developing the selected symbol should not be turned into a claim that IBM supplied the checkout at which the famous gum package was sold. Compatibility between contributions from different companies was part of the system’s value.
Weightman identifies Sharon Buchanan as the cashier and Marsh research and development executive Clyde Dawson as the first shopper. The event was prepared, not an accidental encounter with an unsuspecting customer. In Dawson’s later explanation, the gum package demonstrated that a small product could accommodate a workable code (Weightman, 2015). It was therefore an ordinary type of purchase used to make an extraordinary organisational achievement visible.
That choice had a communicative advantage. An audience can grasp a package of gum more readily than a committee’s technical specification. A familiar object gives an unfamiliar system a scale and setting. Yet the story becomes misleading if the object is allowed to stand for the entire innovation. The scan depended on prior agreement, marked packaging, reading equipment and a functioning checkout process.
The event can be read as a demonstration of confidence in that combination. It made the system legible to people who had not participated in its development. For marketing history, this is an example of implementation becoming communication: practical performance supplied a memorable public story, while the story helped give a complex infrastructure a recognisable beginning.
The Original Gum Package and the Museum Myth
A particularly persistent version of the story places the original gum package in the Smithsonian. A correction appended to Weightman’s article in 2017 explicitly states that the National Museum of American History did not collect it. A facsimile had been displayed to explain the scanner’s history (Weightman, 2015, correction 2017). The distinction is small in physical appearance and substantial in historical meaning.
A reconstruction can explain an event effectively without having been present at it. An original object makes an additional claim through its documented provenance. Confusing the two converts an interpretive device into apparent material evidence. Repetition can make that claim familiar while leaving it unsupported.
This is particularly relevant to the history of marketing, where memorable objects often become shorthand for complicated developments. A product package offers a satisfying centrepiece. It seems to promise a direct connection between a celebrated moment and the present. Historical interpretation has to distinguish that attractive expectation from what an institution actually collected.
The correction does not diminish the importance of the UPC. It makes the account more useful by separating the transaction, its later retelling and the objects used to communicate it. A museum can tell a strong story through an explicitly identified facsimile. What matters is that visitors understand the nature of the evidence. The barcode’s historical significance rests on the documented development and use of a shared system, not on the survival of a particular packet of chewing gum.
Measuring Benefits Without Assuming Instant Profitability
Later ubiquity can make early investment appear self-evidently sensible. Emek Basker’s 2011 study provides a more discriminating picture. Matching scanner-installation information with store-level data from the 1972, 1977 and 1982 US retail censuses, she estimated an early average labour-productivity increase of approximately 4.5 per cent. Short-term gains were nevertheless small relative to fixed costs; her analysis emphasised profitability as an obstacle to diffusion (Basker, 2011).
This helps explain why hesitation need not indicate ignorance. A store can expect operational improvements and still face a difficult investment decision. A gain in output per labour hour is not the same as an immediate return on capital. The timing of expenditure and benefit matters, as does the extent to which an individual store resembles those from which an average effect has been estimated.
The evidence also supports a careful separation of claims. Demonstrating that a scanner reads a product establishes technical performance. Estimating an average productivity effect addresses another question. Showing that a particular installation will be profitable requires further information about its costs and conditions. These statements should not be substituted for one another.
For the history of innovation, that separation is valuable. The UPC’s eventual success does not prove that every earlier decision was easy or that every hesitant retailer misjudged the situation. A shared standard creates a possibility that businesses must evaluate under their own circumstances. The story therefore includes not only invention and persuasion but the ordinary economics of implementation: expenditure now, uncertain benefits later and different outcomes for different types of operation.
A New Location for the Production of Marketing Data
IBM’s account of its 3660 supermarket system, introduced in 1973, describes the connection between product recognition, price retrieval and sales recording. The recorded information could support the management of merchandise (IBM, n.d.c). This establishes an important capability without implying that every early installation immediately used all the analytical applications associated with later retail systems.
The larger change concerns where useful observations could originate. The completed sale itself could become an input to subsequent commercial decisions. Instead of treating checkout exclusively as the end of a purchasing process, a retailer could also treat it as the beginning of a record to be combined with other records. Questions about assortment, timing and performance could then be addressed through information generated during transactions.
That possibility still required organisation. Data have to be linked to the right article, stored consistently and interpreted for a purpose. Reading the printed symbol correctly cannot compensate for a mistaken product record. Nor does a long series of observations decide which comparison matters for a marketing decision.
There is also a limit inherent in the observation itself. A recorded purchase establishes that an item was bought, not why the buyer preferred it or which advertisement mattered. This is a methodological implication of transaction records rather than a criticism of scanning. More systematic observation can support better investigation while leaving explanatory questions open. The UPC belongs to the history of data-based marketing because it helped make those investigations practicable, not because stripes alone could supply complete knowledge of consumers.
Scanner Data and the Scientific Study of Brand Choice
Peter M. Guadagni and John D. C. Little’s 1983 study provides a specific example of what researchers could build on such data. Their brand-choice model used 32 weeks of regular ground coffee purchases by 100 households. Explanatory variables included brand and size loyalty, store promotion and prices (Guadagni and Little, 1983). The important step was the movement from observed purchases to an explicitly specified model of choice.
Sunil Gupta’s 1988 study offered another application. Using scanner panel data for regular ground coffee, it separated the sales increase during a promotion into brand switching, earlier purchasing and stockpiling (Gupta, 1988). A single encouraging sales total could therefore contain different kinds of behaviour with different implications for a brand.
An illustrative example makes the distinction clear. Suppose sales rise during a discount week. Some buyers might have switched from a competitor. Others might have brought forward a purchase they would otherwise have made later. Both appear in the immediate total, but their longer-term significance differs. The example is an explanation of the analytical issue, not a claim about a particular historical campaign.
This is why the significance of scanner data cannot be reduced to collecting more numbers. Their value depends on the questions asked and the assumptions used to answer them. The research history following UPC adoption demonstrates additional intellectual work. Engineers made transactions easier to identify and record; researchers developed ways of distinguishing the processes that could produce the recorded patterns. Marketing knowledge emerged from that combination rather than from automatic measurement alone.
Accuracy Was Also a Matter of Keeping Promises
The checkout remained a place where technical reliability and customer trust could diverge. The 1996 report Price Check, prepared by staff of the Federal Trade Commission, the National Institute of Standards and Technology and participating states, examined whether scanned prices agreed with posted or advertised prices. Its question was not simply whether a scanner recognised the printed symbol (FTC et al., 1996, pp. 1–4).
A product can be correctly identified and still be assigned a price that differs from the shelf label. The report discusses failures to update computer prices promptly and failures to change shelf tags or promotional signs to match. These are organisational problems within a technically functioning system. They concern the coordination of records and physical information throughout a store (FTC et al., 1996, pp. 3–4).
The marketing implications are direct. An advertised offer creates an expectation that has to be fulfilled at checkout. If the charged price differs, the transaction can damage confidence even when the scanner performs its reading task perfectly. Conversely, an informative receipt can help customers check what happened. The quality of the experience therefore depends on the relationship between advertising, shelf information, displays and the final record of sale.
This extends the history beyond the familiar claim that scanning made shopping faster. Speed is valuable, but an offer also needs to be understandable and dependable. The electronic system becomes part of the retailer’s relationship with the customer. Its maintenance is consequently connected to marketing performance, even when the work appears to belong to pricing administration or store operations rather than advertising.
What the 1996 Inspections Actually Found
The joint investigation covered 294 stores across different retail formats, including 113 food stores. It reported an overall pricing error rate of 4.82 per cent and a food-store rate of 3.47 per cent. Both overcharges and undercharges occurred. These are results from that historical study, not estimates for all retailers or present-day checkout systems (FTC et al., 1996, pp. 9–14).
The report also documents an institutional response. In 1995, the National Conference on Weights and Measures adopted an examination procedure for price verification. Common sampling and inspection methods were intended to make pricing accuracy assessable on a more consistent basis. The procedure distinguished the total of pricing errors from overcharges used as the basis for stronger enforcement responses (FTC et al., 1996, pp. 5–10).
That response echoes the original history of the UPC. Agreement was needed first to identify products and later to assess aspects of the transactions built around that identification. A common mark did not remove the need for common procedures elsewhere. Standardisation moved across connected activities rather than ending with the selection of a graphic.
For a historical interpretation, the finding is especially useful because it resists two easy conclusions. The existence of errors does not show that scanning had no value. The usefulness of scanning does not show that continuing oversight was unnecessary. The report describes a working infrastructure whose commercial reliability still depended on people, procedures and maintenance. Its account makes the system more tangible by showing what remained to be done after the celebrated invention had become ordinary.
Maintaining the Connection Between Promotion and Checkout
One example in the 1996 report belongs directly to the history of sales promotion. An offer giving the second item at half price could require manual intervention if the checkout system did not process it automatically. An uninformed cashier could therefore scan both products correctly and still charge the wrong promotional price. The report recommended written procedures, employee training, periodic price audits and an identified pricing coordinator (FTC et al., 1996, pp. 21–22).
The example shows why automation did not remove the need to translate a marketing decision into operating practice. An offer had to be understandable on a sign and executable at checkout. Those were connected requirements, but they involved different work. A breakdown between them could undermine a technically successful scan.
The report also describes handheld scanners for checking shelf prices against stored records, portable printers for replacing labels and trials of electronic shelf tags. It records practical limitations and costs of the latter at that historical moment (FTC et al., 1996, pp. 22–23). Checkout automation was encouraging additional forms of store technology rather than completing the process of change.
The resulting historical lesson concerns maintenance. The UPC’s usefulness depended not only on a successful original agreement but on routines for detecting and correcting discrepancies. These less celebrated activities helped turn the promise of an automated transaction into a dependable retail practice.
Earlier Approaches and International Developments
Printed marks had a machine-reading history beyond grocery retailing. John T. Kermode and Harry P. Sparkes filed their Card sorter patent in 1930; it was granted in 1934. The document describes photoelectric sorting of marked cards and discusses electricity-utility records among its applications. This was not a UPC checkout, but it demonstrates an earlier approach to connecting printed coding with automatic handling (Kermode and Sparkes, 1934).
International development likewise requires its own chronology. The European Article Numbering Association was established in Brussels in 1977. GS1 describes the EAN barcode as compatible with the US UPC. GS1 UK’s timeline identifies a package of Melrose tea bags at Keymarkets in Spalding in 1979 as the first live product scan in its British retail history (GS1, 2025; GS1 UK, n.d.). The American milestone was not a simultaneous global conversion.
The significance of these developments lies partly in compatibility. A national system may work domestically while creating fresh translation problems when goods cross borders. Extending a common method can increase its usefulness without requiring each trading relationship to start again. This is an analytical implication of shared identification, not evidence that international implementation happened without negotiation.
European organisations and local trading structures consequently belong in the story. International diffusion was more than the movement of a finished American idea into otherwise passive markets. It required additional decisions and institutions. Recognising those contributions produces a more accurate global history while preserving the particular importance of Laurer’s design and the US grocery industry’s earlier agreement.
Market Infrastructure and Laurer’s Historical Place
Hans Kjellberg, Johan Hagberg and Franck Cochoy examine barcode scanning as market infrastructure over the period 1967–2010. Their approach connects identification with the arrangements that enable exchange, emphasising relationships, continuing maintenance and political dimensions (Kjellberg, Hagberg and Cochoy, 2019). It offers a useful way to understand why a small package marking belongs in a history of marketing.
The checkout is not an isolated endpoint. It is where agreements made elsewhere acquire practical consequences. A package, a reading device and a product record meet alongside the interests of manufacturers, retailers and customers. The barcode coordinates parts of that encounter without eliminating the separate responsibilities of the participants. Its apparent simplicity conceals the extent of the cooperation on which its usefulness depends.
Laurer’s lasting contribution was to make a crucial technical connection dependable enough for that setting. The 1973 industry agreement gave the design a shared commercial role. The 1974 sale in Troy made its application visible. Subsequent investment decisions, scientific work with scanner data and continuing attention to price accuracy reveal how much followed from, and remained beyond, that initial event.
The history is therefore most informative when its stages remain distinct. A readable symbol was an engineering achievement; a common standard was an organisational achievement; its sustained use was an achievement of implementation and maintenance. Together they changed what companies could record, compare and coordinate in ordinary commerce. The chewing-gum transaction became famous because it gave this larger transformation a memorable scene. George Laurer’s place in marketing history rests on the practical system that made the scene possible.
References
Basker, E. (2011) Raising the Barcode Scanner: Technology and Productivity in the Retail Sector. CES Working Paper 11-16. U.S. Census Bureau. Available at: https://www.census.gov/library/working-papers/2011/adrm/ces-wp-11-16r.html (Accessed: 22 October 2026).
Eastman, J. (n.d.) Brief History of Barcode Scanning. Optica historical collection, pp. 128–133. Available at: https://opticaorgdev.blob.core.windows.net/$web/optica/media/osa.history/century_of_optics/1960-1974/128.pdf (Accessed: 22 October 2026).
Federal Trade Commission (FTC) et al. (1996) Price Check: A Report on the Accuracy of Checkout Scanners. Report by staff of the FTC, NIST and participating states, 22 October. Available at: https://www.ftc.gov/sites/default/files/documents/reports/price-check-report-accuracy-checkout-scanners/scanners.pdf (Accessed: 22 October 2026).
Gerena, C. (2014) ‘Reading Between the Lines’, Econ Focus, second quarter. Federal Reserve Bank of Richmond. Available at: https://www.richmondfed.org/publications/research/econ_focus/2014/q2/economic_history (Accessed: 22 October 2026).
GS1 (2025) GS1 Historical Timeline. Updated 8 July. Available at: https://support.gs1.org/support/solutions/articles/43000734073-gs1-historical-timeline (Accessed: 22 October 2026).
GS1 UK (n.d.) 50 years. Available at: https://www.gs1uk.org/about-us/who-we-are/50years (Accessed: 22 October 2026).
Guadagni, P. M. and Little, J. D. C. (1983) ‘A Logit Model of Brand Choice Calibrated on Scanner Data’, Marketing Science, 2(3), pp. 203–238. https://doi.org/10.1287/mksc.2.3.203.
Gupta, S. (1988) ‘Impact of Sales Promotions on When, What, and How Much to Buy’, Journal of Marketing Research, 25(4), pp. 342–355. https://doi.org/10.1177/002224378802500402.
IBM (n.d.a) George Laurer. Available at: https://www.ibm.com/history/george-laurer (Accessed: 22 October 2026).
IBM (n.d.b) The UPC. Available at: https://www.ibm.com/history/upc (Accessed: 22 October 2026).
IBM (n.d.c) Retail and point-of-sale systems. Available at: https://www.ibm.com/history/point-of-sale (Accessed: 22 October 2026).
Kermode, J. T. and Sparkes, H. P. (1934) Card sorter. US Patent 1,985,035, filed 5 November 1930, granted 18 December 1934. Available at: https://patents.google.com/patent/US1985035A/en (Accessed: 22 October 2026).
Kjellberg, H., Hagberg, J. and Cochoy, F. (2019) ‘Thinking Market Infrastructure: Barcode Scanning in the US Grocery Retail Sector, 1967–2010’, in Kornberger, M. et al. (eds.) Thinking Infrastructures. Research in the Sociology of Organizations, vol. 62. Emerald, pp. 207–232. https://doi.org/10.1108/S0733-558X20190000062013.
Ohio History Connection (n.d.) 35-55 First UPC Barcode Retail Scan. Remarkable Ohio, historical marker. Available at: https://remarkableohio.org/marker/35-55-first-upc-barcode-retail-scan/ (Accessed: 22 October 2026).
Weightman, G. (2015) ‘The History of the Bar Code’, Smithsonian Magazine, 23 September; editorial correction 26 June 2017. Available at: https://www.smithsonianmag.com/innovation/history-bar-code-180956704/ (Accessed: 22 October 2026).
