American Airlines Spent Years Making Its Cabins Worse—Now It’s Racing To Win Passengers Back From United And Delta

United Airlines CEO Scott Kirby says one of his biggest competitive advantages won’t last. Other airlines will catch up with Starlink internet. He says he’s investing anyway, because two or three years is enough time to get passengers to try United, discover other things they like, and move their loyalty and credit card spending. That answers the argument Doug Parker made at American Airlines in 2017 for not investing in product: better products just get copied so can’t offer an enduring advantage. Delta spent nearly two decades showing how investment in the operation, product and employees could work together. United has improved more than any other U.S. airline over the past decade. American has made an impressive amount of progress since its premium pivot roughly 21 months ago. But it’s trying to win back customers while losing money, carrying heavy debt, and with fuel costs up. They’re in a race against time to show results before Wall Street gets impatient. I attended American’s September 2017 media and investor day. The airline’s theory was that competitors would match a better product, leaving corporate culture and employees to differentiate the airline. Take care of employees, who would take care of customers, who would take care of shareholders. (Of course American alienated all three groups.) They didn’t explain how that was going to happen. There were nicer employee facilities coming, help using their health and wellness benefits, and they spent a disproportionate amount of time on company-sponsored weight loss programs. I left without understanding the mission employees were supposed to rally around, or how American would install its workforce with a desire and knowledge to deliver exceptional service. Meanwhile, American was preparing to fly its new domestic interior with less legroom, less padded seats and no seatback screens. At that same media day, they didn’t even think they needed native booking in their mobile app. Sending customers to a mobile website was good enough. A year later, Parker told employees: we work really hard to match our service to our competitors, and they do the same to us. He described watching what competitors put onboard, including meals, to ensure American was comparable. He was surprised by the idea that customers might choose United or Delta because of a different inflight product. In contrast, Richard Anderson became Delta CEO in 2007. Along with Glen Hauenstein and Gil West, he built an airline around keeping its promises to passengers and driving hard bargains with its counterparties. Anderson later explained how much of the advantage came from line maintenance. An older airplane could be reliable if mechanics could work on it where it spent the night, fix recurring problems and have it ready in the morning. Delta also worked toward eliminating involuntary denied boarding. A customer who bought a ticket, arrived on time and followed every instruction shouldn’t be told they weren’t going to travel because the airline had sold too many seats. Finding enough volunteers protected the relationship with that customer. They leaned into seatback screens. Their non-union crews were somewhat friendlier than at competitors. Then Delta made itself relevant in the markets where customers had the most money to spend. The 2011 LaGuardia slot swap with US Airways was one of the great airline asset trades. Parker’s airline gave Delta 132 daily LaGuardia slot pairs in exchange for 42 pairs at Washington National and $66.5 million. Required divestitures reduced Delta’s net gain (they eventually got the slots back that were transferred to WestJet) but it gave them the platform to build the New York business. Delta fleeced Parker. US Airways strengthened Washington, but when it merged with American it had to give up slots there anyway. The combined airline couldn’t keep all of both carriers’ holdings. American – once headquartered in New York – would never be able to position itself in New York the way that Delta had. Delta invested in facilities at LaGuardia, JFK and Los Angeles. That all positioned Delta to overtake American in cobrand credit card spend. I’ve been critical of how much Delta has coasted on that reputation. Its operational advantage isn’t what it was (they’re still a bit more reliable than peers but when things go south they struggle to recover, and no longer go months without a mainline cancel). Their 767s out of New York are arguably the actual worst product across the Atlantic and among those who compete for premium transcon business. And Delta’s success despite weak mileage value shouldn’t have taught competitors to devalue their own programs. Delta offered a better experience – that’s eroding – and had strong positioning in some of the best spend markets, and lack of competition in their other key markets (Atlanta, Upper Midwest, Salt Lake). Kirby took a long time to arrive at his current view. He was president of American when it cut meal service in 2014. Planning for the domestic interior without screens began while he was still there, even though the first aircraft with that interior entered service after he left. He initially carried the same approach to United. They cut full domestic first class meals on flights under four hours outside the dinner window, before customer complaints forced a reversal. They eliminated Polaris wine flights and cocktail carts, then reduced business class staffing by one flight attendant and moved to pre-plated entrees. Some of United’s early product investment was probably overboard – greenlighting most everything when Oscar Munoz came back from medical leave. The later United NEXT commitment to screens, bigger bins and a better domestic cabin, combined with sustained investment in its mobile app, represented a different approach under later Kirby. Most passengers sit in coach. Improving their experience gets them to pay more and builds a reputation that qualifies the airline for their future purchases in business. Even business class passengers internationally are often coach passengers domestically, and bad experiences in domestic form the bulk of a passenger’s opinion when it coms time to buy long haul business. The app is part of the coach product just as well as the premium cabin one. Passengers can manage changes, get help improving their seat assignment and receive clearer explanations when something goes wrong. It gives them more control over their trip. They use it repeatedly, and it’s reason they may hesitate to move their flying elsewhere even after another airline matches the cabin. By 2025, Kirby was explaining a different test for food spending. He told Brian Sumers: Every year we’re going to consciously spend more for customers in a bet that we’ll just win more brand-loyal customers. He couldn’t prove exactly how much additional ticket revenue better food would generate. He was willing to make the investment anyway. The old approach could identify the savings from moving a meal window by 15 minutes, while the revenue lost through a series of small disappointments was much harder to see. United’s network rebuilding started earlier. Kirby restored domestic connectivity that had been cut back under Jeff Smisek, and explicitly explained in October 2017 that better connections would help sell credit cards. If United’s schedule didn’t work for someone, that person wasn’t going to take the card. American went the other way in New York, Chicago and Los Angeles. Retreating made the airline less useful in the country’s most valuable spending markets. Offering a less desirable product alienated customers from spending on the American card. And an airline’s route accounting can miss this. A Chicago customer’s work trips may be what persuade them to get an airline’s card, even if they eventually redeem the miles for Hawaii. Credit the Hawaii flight with all the benefit, cut Chicago because it looks weak, and you lose the spending that generated the miles in the first place. In contrast, aviation watchdog JonNYC shares excerpts attributed to Chief Commercial Officer Andrew Nocella laying out where the airline sees itself: United reached 50% of revenue from Chicago-based passengers in the second quarter of 2026, according to Nocella, and gained share in its other hub cities as well. It expects to largely complete United NEXT’s connectivity goals at its midcontinent hubs by summer 2027. More attention then shifts to coastal gateways and international expansion, including a planned return to New York JFK next year. Premium capacity will grow faster than main cabin capacity as United adds Airbus A321neo and A321XLR aircraft at scale. Comments by Andrew Nocella, shared by JonNYC United’s hubs in some of the best cities for card spend and long haul pre-existed Kirby (and those cities are more contested than other airlines face in their most important hubs). Kirby’s Starlink explanation in his Cheeky Pint interview describes how he plans to accelerate United’s gains. John Collison characterized the opportunity as a two-to-three-year window to win customers, and Kirby agreed. Other airlines will eventually have comparable low Earth orbit connectivity. American, Southwest, Alaska and even Frontier are all adding Starlink (and Southwest expects to complete its install around the time United does). they’re going to try us for Starlink, and they’re going to stay because all the other technology, and everything else is better. A temporary advantage can generate a customer relationship that outlasts it. And the biggest gap will be with Delta, which has committed to Amazon Leo that will maybe start becoming available in 2028, and where they’ve committed only to eventually outfit 500 aircraft. Delta’s current wifi is generally poor now that demands on the system by giving it away free have overwhelmed available bandwidth. Parker was right that competitors could copy a feature. Copying the feature doesn’t on its own win back the customer. American is finally reversing course. I first reported its explicit premium pivot in January 2025. In roughly 21 months they’ve changed an extraordinary amount, from better coffee and wine to improved lounges, new business class suites and more attention to the coach experience. They’ve committed to bringing back domestic seatback screens and adding more first class and extra legroom seats. They’ve announced Starlink installations beginning early next year. They’ve even instructed employees to stop involuntarily bumping passengers, giving airport leadership more authority to secure volunteers. American is changing what it asks employees to prioritize, and giving them tools to do it. And Flagship Dining at Los Angeles is finally reopening after more than six years, with access expanded to qualifying business class passengers. It’s a defensive move because Delta’s business class lounge is so much better, and United’s even had been better than American Flagship. I want this to work. I’ve spent years criticizing American because it had so much potential to be better. They’re making real changes and some customers will come back simply because the airline stops giving them reasons to avoid it. But matching what they already get elsewhere likely isn’t enough – American needs a value proposition for what they deliver to customers that no one else will. That’s Delta’s reliability argument (paired with service and screens that for years others eschewed). That’s United’s Starlink argument (paired with a stronger global network). American entered the pandemic with heavy debt after spending more than $12 billion on share buybacks under Parker. Borrowing while returning that much cash to shareholders left it weaker than competitors. It has too much debt service, less room to borrow, and less well-positioned to do acquisitions. The risk is that investors demand results before enough customers have experienced the improvements, or demand cuts to the spending that could produce real results. American’s domestic seatback screens won’t begin arriving on new aircraft until 2028, and the broader fleet project extends into the next decade. It needs the improvements it can deliver sooner to get customers back onboard and spending more before then. The interest and fuel bills are arriving now. Topics on this page