18.

The Bar Code

There are two ways to tell this story.

One of them describes that classic flash of inventive insight. In 1948, Joseph Woodland, a graduate student at the Drexel Institute in Philadelphia, was pondering a challenge that had been thrown down by a local retailer: Was there any way to speed up the process of checking out in his stores by automating the tedious process of recording the transaction?1

Woodland was a smart young man. He’d worked on the Manhattan Project during World War II. As an undergraduate he’d designed an improved system for playing elevator music. He had planned to launch it commercially but had been warned away by his father, who was convinced that mobsters controlled the elevator music racket.

So Woodland had returned to his studies at the Drexel Institute, and now he was stumped by the transaction problem. On a visit to his grandparents in Florida, he sat on Miami Beach and pondered the problem, idly combed his fingers in a circle, letting the sand slide between his fingers. But as he looked down at the ridges and furrows, a thought struck him. Just as Morse code used dots and dashes to convey a message, he could use thin lines and thick lines to encode information. A zebra-striped bull’s-eye could describe a product and its price in a code that a machine might be able to read.2

The idea was workable, but with the technology of the time it was costly. But as computers advanced and lasers were invented, it became more realistic. The striped-scan system was independently rediscovered and refined several times over the years. In the 1950s, an engineer, David Collins, put thin and thick lines on railway cars so that they could be read automatically by a trackside scanner. In the early 1970s, an IBM engineer, George Laurer, figured out that a rectangle would be more compact than Woodland’s bull’s-eye, and he developed a system that used lasers and computers that were so quick they could process labeled beanbags being hurled over the scanner system. Joseph Woodland’s seaside doodles had become a technological reality.3

But there’s another way to tell the story. It’s just as important—but it’s incredibly boring.

In September 1969, members of the Administrative Systems Committee of the Grocery Manufacturers of America met with their opposite numbers from the National Association of Food Chains. The location was the Carrousel Inn, an uninspiring motel in Cincinnati. The topic was equally uninspiring: whether the food producers of the GMA could come to an agreement with the food retailers of the NAFC about an interindustry product code.

The GMA wanted an eleven-digit code, which would encompass various labeling systems that grocery manufacturers were already using. The NAFC wanted a shorter, seven-digit code, which could be read by simpler and cheaper systems at the checkout. The GMA and the NAFC couldn’t agree, and the meeting broke up in frustration. Years of careful diplomacy—and innumerable committees, subcommittees, and ad hoc committees—were required before, finally, the U.S. grocery industry agreed upon a standard for the Universal Product Code, or UPC.4

Both stories came to fruition in June 1974 at the checkout counter of a Marsh supermarket in Troy, Ohio, when a thirty-one-year-old checkout assistant named Sharon Buchanan scanned a ten-pack of fifty sticks of Wrigley’s Juicy Fruit chewing gum across a laser scanner, automatically registering the price of 67 cents. The gum was sold. The bar code had been born.5

We tend to think of the bar code as a simple piece of cost-cutting technology: it helps supermarkets do their business more efficiently, and so it helps us enjoy lower prices. But like the shipping container, the bar code doesn’t work unless it’s integrated into a system. And like the container shipping system, the bar code system does more than lower costs. It solves problems for some players and creates headaches for others.

That is why the second way of telling the story is as important as the first—because the bar code changes the balance of power in the grocery industry. It’s why all those committee meetings were necessary; it’s why the food retailing industry was able to agree only, in the end, when the technical geeks on the committees were replaced by their bosses’ bosses, the chief executives. The stakes were that high.

Part of the difficulty was getting everyone to move forward on a system that didn’t really work without a critical mass of adopters. It was expensive to install scanners. It was expensive to redesign packaging with bar codes: bear in mind that at the time Miller Brewing Company was still issuing labels for its bottles on a 1908 printing press.6 The retailers didn’t want to install scanners until the manufacturers had put bar codes on their products; the manufacturers didn’t want to put bar codes on their products until the retailers had installed enough scanners.

But over time it became apparent that the bar code was changing the tilt of the playing field in favor of a certain kind of retailer. For a small, family-run convenience store, the bar code scanner was an expensive solution to problems the store didn’t really have. But big-box supermarkets could spread the cost of the scanners across many more sales. Such stores value shorter lines at the checkout. They need to keep track of inventory. With a manual checkout, a shop assistant might charge a customer for a product, then slip the cash into her pocket without registering the sale. With a bar code and scanner system, such behavior can be pretty conspicuous. And in the 1970s, a time of high inflation in America, bar codes let supermarkets change the price of products by sticking a new price tag on the shelf rather than on each item.

It’s hardly surprising that as the bar code spread across retailing in the 1970s and 1980s, big-box retailers also expanded. The scanner data underpinned customer databases and loyalty cards. By tracking and automating inventory, it made just-in-time deliveries more attractive and lowered the cost of having a wide variety of products. Shops in general, and supermarkets in particular, started to generalize—selling flowers, clothes, and electronic products. Running a huge, diversified, logistically complex operation was all so much easier in the world of the bar code.7

Perhaps the ultimate expression of that fact came in 1988—when the discount department store Walmart decided to start selling food. It is now the largest grocery chain in America—and by far the largest general retailer on the planet, about as large as its five closest rivals combined.8 Walmart was an early adopter of the bar code and has continued to invest in cutting-edge computer-driven logistics and inventory management.9

Walmart is now a major gateway between Chinese manufacturers and American consumers. Walmart’s embrace of technology helped it grow to a vast scale, and its vast scale means that it can send buyers to China and commission cheap products in bulk. From a Chinese manufacturer’s perspective, it can justify setting up an entire production line for just one customer—as long as that customer is Walmart.10

Geeks rightly celebrate Joseph Woodland’s moment of inspiration as he pulled his fingers through the sand—or the perspiration of IBM’s George Laurer as he perfected the bar code as we know it. But the bar code isn’t just a way to do business more efficiently; it also changes what kind of business can be efficient.

Yes, those distinctive black and white stripes are a neat little piece of engineering. But that neat little piece of engineering has changed how the world economy fits together.