The Toll of a New Machine

Melnik, 39, is one of those smart, restless souls who stumble into entrepreneurship because it makes so much more sense than working for companies that are too big, too slow, and too hesitant. A college dropout whose real passion is marine biology (for a while, he performed as "Flipper Boy," cavorting with dolphins at the Miami Seaquarium), he started the company that became Kinetics in 1988 after working as a sales agent for a Tampa travel agency. The experience of selling and delivering airline tickets to small businesses got him wondering why those tickets couldn't be handled more like money in ATMs and less like a special product that had to be "produced" at a travel agency. "I got interested in this, and it hooked me," he says.

Melnik worked with NCR on several projects, including plans to bring an early kiosk to the Trump Shuttle in New York's LaGuardia Airport. He worked with Siemens on a project for Lufthansa. He also worked as a waiter, accepted startup funding from his mother-in-law, and lived for several years off the teacher's salary of his wife, Cindy. He can write software, and he can "bend metal" to make kiosks.

It took nearly eight years of persistence for the technology, the airline mind-set, and the customers to catch up with Melnik's vision. He sold Alaska Airlines on those first machines, called Orcas, using a cardboard mock-up.

How this kind of simple but powerful self-service technology ripples through businesses and the economy always looks easier to predict than it is. The first passenger elevator in the world, created by Elisha Otis, was installed in a New York department store in 1857; it wasn't until 1950--nearly a century later--that the Otis Elevator Co. came up with the technology for self-service elevators. In 1955, 500,000 people in the United States were working as elevator operators, jobs that were almost all gone less than 10 years later.

But as it turns out, the impact of even the most pervasive self-service, on productivity and on customers, is easy to misinterpret. Kinetics' machines improve the productivity of airline ticket agents--but not by allowing the ticket agents to do more work, more quickly. They allow the ticket agents to preside over more work being done--in this case, by the customers. And it may be this sort of productivity improvement that helps make possible the "jobless recovery," in which companies manage to grow without hiring new employees or without recalling those who have been laid off.

Indeed, when you use a self-check-in machine, you can't help but wonder about the thousands of airline employees who have lost their jobs since September 11. Last year, Northwest flew 12% fewer passengers than in 2000. But it did so with 25% fewer employees. If Delta had been staffed in 2003 the way it was in 2000, it would have employed 2,500 more people. Since the end of 2000, Kinetics' three biggest customers--Delta, Northwest, and Continental--have shed some 37,000 employees, enough people to run all of Northwest today.

Of course, the airlines are a complicated case--their business was out of whack before the September 11 attacks, and the attacks hit the airline business harder than any other. But even where the impact of such machines looks obvious, labor tends to squirt around the economy in unpredictable, even counterintuitive, ways. Although Continental now has 780 kiosks in 130 airports, with the machines handling the vast majority of passengers, the airline has reduced the number of airport agents by only 4% since September 11.

Melnik likes to say that each Kinetics self-check-in machine, at an initial cost of between $6,000 and $10,000, takes the place of two-and-a-half ticket agents, because the machine is available (at least) from 6 a.m. to 9 p.m., seven days a week--or about the number of hours that many agents would work. But that both understates and overstates the machines' impact.

Kinetics has installed 3,800 self-check-in machines for airlines--but 9,500 ticket agents have not lost their jobs. At the same time, at airports in Atlanta and Houston, where there are banks of dozens of check-in machines, the kiosks handle surges of passengers easily and quickly. No airline can have 50 or 100 ticket agents waiting to take bags and issue boarding passes; but many airlines have that many check-in terminals in individual airports.

At Continental, O'Leary acknowledges that the airline is using Kinetics' technology to grow traffic without adding staff and costs. "It's absolutely true that before self-service, we were adding staff and [airport] real estate like you wouldn't believe," he says. "Once you have self-service deployed, you can absolutely contain those costs. But we still argue we are getting better productivity and service out of our existing agents."