Why Boeing shouldn’t pull the plug on Connexion.

The Wall Street Journal reported Thursday that Boeing, the giant aerospace company, is thinking about either closing or selling its Connexion by Boeing unit, which offers in-flight Internet service.

At first blush, the story looks like another case study of an expensive, hype-ridden technology infrastructure venture that failed to pan out. Boeing, the Journal reports, has spent $1 billion rolling out the system. But because travelers have been slow to shell out for the service, the unit is only worth $150 million. Dig a little deeper, however, and Boeing’s potential move to unplug Connexion may wind up as a neat case study of a corporation throwing in the towel too early.

Connexion was conceived in April 2000, at the height of the Internet bubble. At the time, Boeing proclaimed the service would “revolutionize the way people travel” and threw about the pro forma math common at the time. The release noted that “analysts project the program’s addressable market to be about $70 billion over the next 10 years.”

As with many of the infrastructure-heavy businesses conceived in the late bubble—Global Crossing, Webvan, eToys—management proved to be wildly optimistic about the rate at which service would be rolled out and the rate at which consumers would take to the new service. In the spring of 2000, Boeing promised system installations would start in late 2001, “with operability expected shortly thereafter.” To hasten the process and give airlines a stake in the new technology, in the spring of 2001, Boeing signed a letter of intent to create a joint venture with American, Delta, and United—three airlines that carried lots of business customers on long-haul flights—to install the service on 1,500 planes, starting in 2002. And in June 2001, Lufthansa committed to be the first foreign carrier to use the service.

The rest is aviation history. The attacks of Sept. 11 and the ensuing airline mass bankruptcies brought the plans to a screeching halt. The joint venture never got off the ground. Delta and United filed for Chapter 11, and other U.S. airlines went into a defensive, cost-cutting, capacity-reducing survival mode. No U.S. carrier was willing to commit the $500,000 per plane needed to install the equipment.

And as is frequently the case with communications systems, the roll-out was slower than expected. In July 2002, Japan Airlines signed up as the first Asian user, and European and Asian airlines signed up through 2003. But the service didn’t formally premiere until May 2004 on Lufthansa—more than two years after originally promised. Throughout 2004 and 2005, Connexion gained traction, signing up European, Asian, and Middle Eastern airlines, including Israel’s El Al and Etihad, the national airline of the United Arab Emirates. (Here’s the most recent list of airlines that offer the service.) In January 2006, Connexion reduced its prices: It now costs $27 to use the service for 24 hours and $18 for three hours.

The technology clearly works and is great value. I tried it on a trip to Germany in the fall of 2005, and it was excellent: reliable and fast. Stuck in economy class for eight hours, I was able to keep up on news, annoy friends (Hey, I’m on the plane! And I’m sending you an e-mail!), reassure my wife that the flight was on schedule, and even do a little work. And since disembarking, I’ve added Connexion to that list of things that are readily available abroad—good German beer, fresh French baguettes, ultra high-speed Internet service in Korea—but that are frustratingly unavailable at home.

Indeed, it remains something of a mystery as to why no U.S. airline has signed up for Connexion. After all, American travelers are desperate for ubiquitous Internet access. Professionals finger their BlackBerrys like worry beads on trains, in airport lounges, on street corners. Wherever possible, they fire up their laptops to get online—in Starbucks, in airport lounges, in New York’s Bryant Park. Train commuters have taken to installing cellular Internet access on their laptops. But planes are the last Internet-free zone. And because we inhabit a social and professional world in which it is no longer acceptable to be completely out of pocket, a kind of panic sets in when the pilot issues the command to turn off electronic communication devices.

Time is money for every knowledge worker—whether unpublished novelist or investment banker. And I’m sure a great number of what Robert Reich calls symbolic-analytic workers would gladly pay to avoid having to watch a terrible movie or talk to a stranger and instead spend their flight time reading Slate, shopping, working, or checking the weather at their destination.

Sure, Connexion costs money—it costs money for the airlines to install it, and it costs money for travelers to use it. But for all parties, it would be money well-spent. Think of the competitive advantage JetBlue has reaped by putting seat-back televisions on its planes. Airlines that cater mostly to business customers could certainly do the same by making their planes Internet-ready. When a company sends a person away on business, it is already shelling out thousands of dollars—full freight for business class, car services, hotels, and meals. An extra $25 for Internet service is a drop in the bucket.

If it was really serious about proving the utility of its business model, Boeing shouldn’t rely on the sick airlines. The company should front the $500,000 per plane to outfit a bunch of planes that fly routes where there would be a lot of use—the Boston-New York-Washington shuttle, the West Coast, and Dallas-based long-haul flights—and see how many people are willing to pay. If, say, 5 percent of the 730 million passengers who fly U.S. airlines each year were to pay an average of $20 for in-flight Internet access, that would be $733 million in annual revenues.

By holding on to Connexion, Boeing could also break the long and repeated cycle whereby the people who lay down the crucial new commercial infrastructure—rails, telegraph lines, fiber-optic cable, warehouses for online grocers—wind up losing their shirts, and the people who buy them at fire-sale prices make out like bandits. High-flying venture-backed companies like eToys and Webvan and Global Crossing failed in the last boom because they lacked the capital to tide them over until the market came around. Boeing doesn’t have that problem.

If Connexion is shuttered or sold at a low price, it will be tempting to view it as evidence that in-flight Internet access was an idea whose time has not yet come. Instead, it’s an idea that hasn’t really been tried.