At a typical consumer-facing company, members of the product team might be the all stars. They design cool gadgets with the newest technology, and if their creations sell well, the company succeeds.

Product teams are important at airlines, too. They design new seats and lounges, scour the world for food and wine, and source soft goods like pillows and blankets.

But they don’t necessarily make a carrier tick. Instead, an airline’s rockstars are often network planning executives, who decide where the carrier will fly, how often, at what times, and with what aircraft. They determine whether you’ll have a 45-minute connection in Charlotte, or a three-hour one. They make sure planes are in the air as much as possible — the only place where they make money.

From a customer angle, it should make sense. Travelers, especially lucrative business customers, want efficient routings, and prefer to fly at the most convenient times. When they connect, most want to spend as little time at the airport as possible, no matter how opulent the lounge. They won’t fly out of their way, or book a six-hour connection, for marginally more comfort to Tokyo. Even price-sensitive travelers value efficiency, often preferring their closest airport, rather than one an hour away, if the price is right.

In short, no carrier will succeed if it has the world’s best product but doesn’t fly where its customers want to go.

At American Airlines, the world’s largest carrier, Vasu Raja, is American Airlines’ vice president for planning, leading the team that schedules American’s roughly 6,500 daily flights. In recent months, he has added a slew of new cities, from South Bend, Indiana to Venice, Italy.

We recently met with Raja in his Fort Worth, Texas office. Here are five takeaways from the conversation — four serious, and one fun.

Ultra Long-Haul Is Not a Priority

In September, about seven months after Patrick Quayle left Raja’s team to run United Airlines’ international network, Quayle had some fiery words about his former employer, while announcing United’s new Houston-Sydney route, one of the world’s longest.

“The real story here is that United is successful at ultra-long-haul flying whereas American and Delta are not,” Quayle told Forbes, adding “It’s easy to add a Venice flight in the summer and make it work,” but tougher to fly 17 or 18 hours and make money.

Raja declined to speak specifically about United, which has added several new ultra-long routes, including Los Angeles to Singapore. But for American, Raja said the economics of ultra-long-haul can be tricky. American has only one true ultra-long-haul route, Dallas to Hong Kong, about 17 hours, though it has a few in the 15-hour range.

“It’s all about the returns you get on the airplane,” Raja said. “If you go and fly one Pacific route, on balance, that’s two airplanes. Are those two airplanes better in the Pacific or are they better flying in the Atlantic for six months of the year and flying to Hawaii for six months of the year?”

American prefers to fly the highest margin routes, rather than the longest, he said.

“In many of these far-flung destinations, you’ve got to tie up an asset or multiple assets for a long time,” he said. “But very often the total [air] market size is smaller than Lexington, Kentucky. It’s just a really risky proposition. It is not something we’re averse to but, in general, when we go into it, we like to have a partner on the other end or some knowledge of the market that helps go and de-risk it.”

Eventually, Raja said, routes to Chinese cities other than Shanghai and Beijing might work. In the past five years, before Quayle joined, United added Hangzhou, Chengdu and Xi’an, but today, only Chengdu remains, as the others performed poorly.

“We look at interior China a lot,” Raja said. “But there are interior China markets that are growing at 35 percent and 40 percent in terms of their actual absolute airline bookings, but that makes them probably smaller then Del Rio, Texas.”

American Still Flies Some Unprofitable Routes

Once, many U.S airlines lost money on a substantial number of routes, keeping them because they hoped they would improve, or because they were crucial to the overall network. A global airline probably must fly to London, whether the flight makes money or not.

But today, Raja said, American has few unprofitable routes.

“This is something often called strategic flying — a euphemism for losing money but telling yourself it’s OK,” Raja said. “We have a really tight and almost dogmatic definition of what strategic flying is: If we’re flying something that doesn’t earn the returns of the rest of our system, either it needs to be on a path where it can earn those returns, or we need to be able to prove pretty definitively that flying it is crucial to [high returns] elsewhere.”

Raja said American has “very few flights” that don’t make money, and most of the losers are on a path to profit. Raja said he expects that includes Los Angeles-Beijing, launched in November. It’s a fiercely competitive route with low fares, but American is hopeful its partnership with China Southern Airlines will boost it. 

“We started it deep in its booking curve and it didn’t do particularly well,” he said. “But over the long run, Los Angeles is going to be our Asia-Pacific gateway. All the other international flights at L.A. have improved by double digit margins over the last two years. We recognize LAX to Beijing probably won’t be that great in January or February, but in January and February of 2020 we expect it to be pretty good, and the more we can integrate with China Southern, the better it gets.”

American has experience with strategic flying in Los Angeles. About a decade ago, American began focusing on Los Angeles International, and it’s now No.1 in market share.

“When we started Los Angeles several years ago, it was unprofitable,” he said. “Los Angeles today is profitable. We anticipate by the end of the year it probably will earn profits in line with the majority of our hubs. If we had canceled markets in L.A. — years ago we could have canceled Los Angeles to Atlanta or Los Angeles to Orlando because it wasn’t doing particularly well — that would’ve been bad news.”

American Has Not Given Up in New York

New York is different. For decades, American owned the market, shuttling high-value customers across the United States, Europe, and to Japan, but the environment has changed, as its competitors have grown, and American has other priorities.

Delta Air Lines now has as many as 275 daily departures from LaGuardia and 230 from JFK. Meanwhile United, after absorbing Continental Airlines’ hub, has 400 flights at Newark, and robust service to Chicago, Houston, Denver and Washington, D.C. from LaGuardia. Plus, JetBlue Airways flies up to 225 daily departures from New York-area airports.

American has roughly 170 departures at LaGuardia and 80 from Kennedy, a spokesman said. From JFK, it no longer flies to Asia — its Tokyo route, ended in 2013 was a disaster — and it has cut many Europe routes, focusing on London and Spain. For American, many European markets work best from Philadelphia, a hub the airline inherited from US Airways with more scale and better connectivity.

“We know who we are in New York,” Raja said. “Our value proposition is to be the best airline for the business customer, operating killer schedules in all the biggest business markets. But that’s what our New York is, and increasingly that’s what we’re focusing it into.”

Top performers include shuttles from LaGuardia to Boston, Washington, D.C. and Chicago, as well flights from Kennedy to London, Los Angeles, San Francisco and some other West Coast cities. But other routes, like New York to Manchester, UK and Zurich no longer make sense, Raja said.

“Kennedy to Seattle, San Diego, Portland, Phoenix — any one of which are gigantically bigger business markets than Kennedy to Manchester are, and all of which are ones which we could do better in for that reason on a P&L basis,” Raju said. “In New York, our core corporate markets generate profit margins, in line with any of our core hubs.”

Flight Numbering is Random

Raja and his team also decide flight numbers —  but it’s not as much fun as it once was.

“Network planning departments of yore are filled with stories about how the VP of Network Planning selected the flight numbers — asking what would be flight No. 1 and flight 666,” he said. “Today there’s a handful of flights —mostly major international flights — in which the flight numbers are locked. For example, Dallas-Heathrow is flight 50. But, other than that, no human numbers our flights. We have a machine that numbers all the flights.”

Before American switched to automation, Raja said it took humans a week or two to number flights for every schedule. “Keep in mind, we make 11 major schedules a year,” he said “Effectively, 11 to 20 weeks per year was spent flight numbering the airline.”

The system works with parameters, so there’s no Flight 666, and American generally doesn’t use numbers associated with crashes. It permits numbers in the sometimes-unlucky 1300-range, unlike United. Otherwise, though, it’s mostly random.

“Numbering is completely disconnected from whatever its probably original intent was,” he said. “A great example is the DFW-Heathrow flight. That had some significance to AMR — [Americans’ predecessor company, which disappeared with the merger] but whatever it is, nobody in the modern company has the history of why it was flight 50 and 51.”

Raju kept referring to DFW-London for a reason. After the merger, American let the machine re-number it. But “the customer outcry was significant,” so American returned to 50 and 51.

For the same reason, Flight 1 remains New York to Los Angeles.

Want to know about network planning from different perspective? In December, we published an interview with Kristen Schilling-Gonzales, director of planning at Allegiant Air, a U.S. discounter. Learn why Allegiant doesn’t send many MD80s to Las Vegas during summer.

Photo Credit: An American Airlines Airbus A321 departs New York JFK for Boston in 2016. American is content with its strategy in New York, an executive said. Alan Wilson / Flickr