The grocery industry settled on one bar code in 1973 and a pack of gum rang it up first

In the 1960s American grocers had a profit problem the cash register couldn't fix: checkout queues moved at typing speed and stocktaking meant hand-counting shelves. Kroger's 1966 booklet ended with a public plea — 'could an optical scanner read the price and total the sale.' RCA answered first, building on the 1952 bull's-eye patent; by July 1972 its automated checkstands were live in a Cincinnati Kroger. But one working system in one chain was worthless to an industry of thousands of stores and hundreds of thousands of products: the code had to be universal before any of it made sense.

So the grocery trade created the Ad Hoc Committee of the Universal Product Identification Code to pick one symbol for all groceries. It set the constraints before judging candidates: fit in 1.5 square inches, print on existing label technology, read in any direction at speed, keep undetected errors below one in 20,000. RCA pushed its proven system. IBM entered late with no product, and engineer George Laurer built a rectangular bar code to the spec instead. On March 30, 1973, after an MIT appraisal, the committee chose Laurer's design: long bars survive press smear; circles do not.

On June 26, 1974 at 8:01 a.m., at a Marsh Supermarket in Troy, Ohio, cashier Sharon Buchanan scanned a 10-pack of Wrigley's Juicy Fruit chosen by Marsh's R&D chief precisely because nobody was sure such a small package could carry a code. The NCR register rang up 67 cents. Adoption was slow at first — a $4,000 scanner and a $10,000 checkstand took longer to pay back than promoters promised — but the network effect took hold: the more products carried the code, the more every scan saved, until mass merchandisers like Kmart and then Walmart made it the backbone of retail.

One open code; print it once, scan it everywhere

A public standard broke the chicken-and-egg trap: makers would print a code only once stores could scan it, and stores would buy scanners only once everything was coded.

Printability on existing presses turned half a million SKUs' worth of packaging into free infrastructure — the code cost manufacturers almost nothing to adopt.

Robustness beat incumbency: RCA had the working system, but the linear symbol's tolerance of ink smear and upside-down reading matched the reality of high-speed printing and lazy checkout clerks.

Decoupling code from hardware kept the layer competitive — any vendor could build scanners, and NCR, IBM and others drove the price from $4,000 toward a rounding error.

Write the spec first and let vendors fight inside it. Pick standards for real-world mess, not demo elegance: presses smear ink, and a long bar tolerates a smear that a circle cannot.

The barcode took off in the 1980s as scanner prices collapsed; Kmart led among mass merchandisers and Walmart built checkout-data logistics on top of it. By 2004 Fortune estimated 80 to 90 percent of the top 500 US companies used the barcode. The Uniform Product Code Council became GS1 US, and the UPC's 12 digits grew into the global GTIN system — one committee's spec now silently prices and tracks most goods sold on Earth. Every loyalty program, warehouse scan and 'what sells where' report descends from that 67-cent pack of gum.