Exposed! Your Torrid Love Affair With Airline Fees

In this issue:

• Exposed! Your Torrid Love Affair With Airline Fees—Airlines offer a commodity product in the very long run and in the very short term, with monopolistic traits in between. This means that the economics of price discrimination and annoying fees are counterintuitive: in the long run, they increase the return on any given plane, but for airlines that just leads to higher orders. The net impact is that, however annoying junk fees are, they actually make air travel cheaper, and they work because, when given the option, a set of consumers will opt for the superficially cheap fare over the more elaborate bundle.
• Malicious Compliance—An EU-mandated Google update makes Google more anticompetitive (at least according to a competitor).
• Sanctions and Third Countries—Non-aligned countries are willing to do business with both sanctioned countries and countries doing the sanctions.
• Bing and BATNAs—An ideal world for Apple is one where Bing is always close to getting acquired by them, or beating Google for the search deal, but never quite gets there.
• Celebrity Endorsements and Binary Outcomes—The economics of selling a product that can only get ~0% or ~100% adoption in its target market.
• Supply/Demand Balance—The GPU shortage slightly abates, but for now this doesn't affect Nvidia's numbers at all.

Exposed! Your Torrid Love Affair With Airline Fees

Markets are mirrors. Look at them, and what's reflected back is the aggregated, dollar-weighted preferences of everyone with similar desires, adjusted for what real-world constraints will allow. There are times when this is frustrating because the average person doesn't share your preferences, but, on average, you’re just another component of someone else's "the average person," and your market-affected preferences are about as annoying to them as theirs are to you.

Airline fees are a classic example of something people complain about but secretly want. This is not a "want" in the direct sense, where anyone has the explicit goal of paying extra for checked bags. But it's true in the indirect sense: given the opportunity, almost everyone will almost always opt for choices that make it a better business decision for airlines to unbundle different aspects of a flight and then charge for them separately.

This has made airline fees a live political issue, as well as a general annoyance. And it's true that it's very annoying to go through a checkout process where prices are prominently displayed—and then find out that every single one of these prices was wrong (and no, they weren't too high). That is not an ideal user experience, but it's also the result of understandable behaviors on the part of airlines, online travel agencies, and customers.

The broad context for understanding this is that airlines are a famously, notoriously cyclical business, one whose cycles are so extreme that the industry has on occasion been net unprofitable for its entire existence. But that cycle buries three different cycles that move at different paces:

1. The obvious cycle is that airlines' assets are mostly planes, and these take a long time to build. Boeing, for example, delivered 528 aircraft last year, and has a backlog of 5,626, or 10.7x annual shipments. Because of the lag, the supply of aircraft always peaks after demand peaks, which crushes pricing; companies can and do shuffle orders around, so this backlog isn't entirely real, but it does reflect the fact that there are almost always a bit too few planes in the air, unless there are far too many on the ground. Meanwhile, one of their biggest costs is fuel, which has its own cyclical dynamics but tends to have big moves in the wrong direction.^[1]
2. A smaller epicycle that sometimes drives this has to do with the level of growth opportunities in the air travel market, and financial markets' appetite for funding incumbents. A budget carrier can deliver good initial numbers because it'll start with new planes and new pilots, and thus a more favorable cost structure. Those costs get worse every year since fleet age predicts the number of repairs needed and therefore the optimal size of their spare parts buffer, while employee tenure predicts salary. Fortunately for airlines, some of their costs go down as they expand, so a budget carrier can maintain that relative cost advantage through breakneck expansion.
3. Airlines vary, by place and time, in how much political pull they have. In many jurisdictions, economic resources get shifted from other sectors towards airlines because they offer bragging rights. The US did this early on with generous mail delivery contracts and strict price regulation, but has since loosened things up a bit, leading to materially lower prices for consumers.

On point three, it's worth digging into exactly why those prices got so low. One reason, from the article linked above: "[L]oad factor, or the percentage of seats that were filled with paying passengers, fell from 66% in 1951 to 53% in 1968. With regulated airfares, it was impossible to slash prices to fill seats, and emptier planes raised per-passenger costs." This is still true today: overall, US airlines had a load factor in November of last year of 81.3%; Ryanair was at 92%. (In the summer, they can get up to the high 90s.) Full planes mean cheap seats. And price discrimination is what fills seats. Are you desperate to get from point A to point B? Are you willing to sit in a tiny seat with no legroom, risking the possibility that your seatmate will literally overflow into your spot?

For a while, the network carriers—American, United, and Delta—resisted competing directly with low-cost carriers on price. They argued that the experience was simply better than what someone like Spirit or Allegiant could offer, and that it would hurt their branding if they offered a stripped-down experience. The problem they ran into was that most people search for flights online, and a large number of these people look at the "price" column and not the "airline" column. A good way to win a sort-by-price contest is to take features out of the product, strip them out of the price, and then sell them back to the customer after they've already made their purchasing decision. If a free checked bag is costly, you can drop it from the bundle, cut the price, and expect to make that revenue back since many customers will, in fact, check a bag.

But not all of them! So Basic Economy seating, while hated, was in fact a price cut for some customers, specifically the ones who didn't value any part of the ticket bundle other than the transportation. Meanwhile, people who need to check bags, or who may need to rearrange their travel plans, pay up. It's possible that the all-in bundle is more expensive than an economy-class ticket used to be, but that's not easy to maintain:

1. Basic economy is an offering that makes sense in markets where network carriers compete with low-cost carriers, so there's already going to be competitive pressure.
2. On routes where this competition isn't a factor, it already wasn't before. So for the new bundle to be too expensive, airlines have to have been undercharging before the rise of basic economy fares.

This pricing structure has benefits, even for people who don't take advantage of it. Every passenger on the plane is defraying some fixed costs, so full planes can lower prices. And the way to optimally fill a plane is to charge a minimal cost for the last seat that would otherwise fly empty, but to only offer that price to someone for whom the price is the deciding factor in whether or not to travel. A typical flight will have some people who absolutely needed a particular departure/arrival window; if you've got a breakfast meeting in San Francisco and a dinner to get to in New York, your options are limited; often, many of the people with such schedules find these things added to their calendar at the last minute, so the airlines charge higher prices closer to departure. Some travelers are loyal to mileage plans first and airlines second. So there are lots of opportunities to put different offers in front of different people, such that a substantially equivalent product can be sold at many different price points. One way to think of this is that there's a platonic price for an economy-class ticket on a given flight, and there's a curve representing the premium some passengers pay and the discount that other ones get. And the steeper that curve is, the lower the midpoint can afford to be while still making the flight profitable. If there were a way to sell the last few tickets on a nearly-full flight at 80% off, without affecting future consumer behavior, not only would this make those specific tickets cheaper for the passengers who bought them, but it would lower the breakeven price for every other kind of ticket.

American consumer spending runs at almost $16tr annualized. It's clear that Americans are willing to spend. But getting them to spend some of that money on your particular business is about the most competitive field on earth. There are countless tricks for doing this, everything from "just make the best product" to "just hire the best PR agency" or even the Ryanair and early Walmart tactic of making the product so absurdly cheap that the cheapness is itself newsworthy—there is a magical point at which you can give up a point of gross margin and get the press and word-of-mouth equivalent of spending more than 1% of revenue on sales and marketing.^[2] In general, it takes the deployment of a large number of techniques to get people to actually open their wallets and spend, and it's no coincidence that one synonym for "techniques" is "tricks," like hiding the final price behind some more favorable-looking base price that omits many costs customers ultimately pay.

Ironically, regulating junk fees follows exactly the same economic logic as imposing them in the first place. The result will be a superficially better-looking shopping experience that turns out to be more expensive in the aggregate and marginally less convenient. Because the airline industry is broadly commoditized, the immediate hit to airlines' economics will be temporary, and they'll all adjust. But in the long run, reducing airlines' ability to charge everyone exactly what they're willing to pay reduces the present value of the next plane they consider buying, meaning that there's ultimately less travel at higher prices. It might be a fair tradeoff to reduce the aesthetic revulsion of fake prices and a multi-tier system even if the fully transparent and egalitarian version provides a little less of the end product.

Diff Jobs

Companies in the Diff network are actively looking for talent. See a sampling of current open roles below:

• A systematic hedge fund is looking for portfolio managers who have experience using alternative data to develop systematic strategies (NYC).
• A company building the new pension of the 21st century and building universal basic capital is looking for a product designer with fintech experience. (NYC)
• An investment company using AI to accelerate investment in esoteric asset classes is looking for a product engineer with Python and Typescript experience; preferably someone with a track record of building on their own (Bay Area, remote also a possibility).
• A crypto proprietary trading firm is actively seeking systematic-oriented traders with crypto experience—ideally someone with experience across a variety of exchanges and tokens. (Remote)
• A concentrated crossover fund is looking for an experienced full stack software engineer to help develop and maintain internal applications to improve investment decision-making and external applications to enable portfolio companies. (SF)

Even if you don't see an exact match for your skills and interests right now, we're happy to talk early so we can let you know if a good opportunity comes up.

If you’re at a company that's looking for talent, we should talk! Diff Jobs works with companies across fintech, hard tech, consumer software, enterprise software, and other areas—any company where finding unusually effective people is a top priority.

Elsewhere

Malicious Compliance

The Digital Markets Act mandates that large platforms in the EU not use their dominance in one area (like search) to funnel business to something else they do (like local business reviews). Google is testing tweaks to its layout to accommodate this, and Yelp argues that, based on user testing of the new interface, these changes will actually increase use of Google's own services. One of the paradoxes of tech regulation is that the industry itself struggles to determine which things are standalone products and which are actually features that will be incorporated into every product by default. Is a dictionary something consumers should pay for, or is it a basic expectation that your word processor knows that "tpyo" is a typo?

This is particularly hard in search, because the problem it solves is so general. Users want answers to questions, but they also need some prodding to come up with the right questions. And the ideal "answer" isn't necessarily a link to a webpage; sometimes it's just the answer, and sometimes users want the most convenient possible good-enough summary of the information they're seeking. It's easy to say from the outside that Google is kneecapping Yelp, but it's entirely possible that from Google's perspective, Yelp is a company that misjudged the optimal bundle and is now using regulation to provide the economic viability that its underlying economics don't offer.

Further complicating this is how path-dependent the whole process is. Google is keenly aware of disruption risk, and knows that any time it controls distribution but someone else controls the underlying content, it's at risk of losing if customers go direct. So Google has, in the past, been more aggressive at thwarting smaller companies early in their life. Some behavior that cut query volume in a given category by 1% last year and 2% this year is a big problem for Google if next year it's 4% and the year after that the category is ex-growth for Google. So when they react, they react early. And because they're extrapolating growth trends, not just present behavior, they tend to be more aggressive in responding to small challengers who are growing fast, i.e. the more users love a Google alternative, the smaller that alternative can be before Google tries to squash it. All of this sets up some truly perverse tradeoffs for Google: the bigger a threat to Google some company is, the more sympathetic it is as a Google victim.

Sanctions and Third Countries

In theory, one of the most powerful policy tools the US has is its control over the dollar—what kind of interest you can earn on it, where you can send it, and, in some cases, whether or not the dollars you've earned are available to you at all. But this monopoly is not absolute; The Economist looks at how non-aligned countries are doing thriving business with sanctioned countries like Russia and Iran, sometimes importing sanctioned oil and then exporting non-sanctioned oil-based products, including to the US itself ($). The goal of sanctions is to make some product less fungible; there aren't just physical grades of crude, but legal grades, too. But this only works in the long run with a monopoly. The US can and does extend secondary sanctions, punishing companies for doing business with sanctioned entities. But this is a risky proposition, since it means punishing a company for behavior that was legal in every jurisdiction involved. Making a habit of this means weakening the dollar ubiquity that makes sanctions work in the first place.

Bing and BATNAs

Google says that Microsoft tried to sell Bing to Apple in 2018, but couldn't get the deal done because Apple didn't like the quality of Bing's search engine. There is a continuous undercurrent of deals-in-some-vague-form-of-progress in big tech, partly because the circumstances in which a company might be willing to part with some strategically interesting asset can change so much. And the diligence process is a form of intelligence-gathering, on both sides.

One thing worth considering here is how much more valuable Bing is to Apple if it's outside of Apple, particularly if rumors of an acquisition float around from time to time but the details of why it didn't happen are kept secret. Bing's existence means that Apple has a reserve bidder for default search placement, which means that Google won't walk off with all of the economic profits. If Apple had an in-house search engine, whether from Bing or one built internally, they'd have one real shot at replacing Google, after which their financial results would make it obvious to Google whether this had worked or not. Whereas if Bing is always in the running but never quite gets there, Google can't be too aggressive in how much Safari search revenue it collects. And if Apple indicates to Microsoft that Bing isn't an asset they want to own because of specific problems, it nudges Microsoft to fix those problems instead of leaving search entirely.

Disclosure: Long MSFT.

Celebrity Endorsements and Binary Outcomes

One of the whistleblowers at Wirecard is launching a startup that makes it easier for employees to report wrongdoing at their employers. Ideally, this would be standard, and every company would have a default mechanism for low-level employees to skip a few levels in the org chart in order to warn senior executives or the board of directors about bad behavior. But there isn't a non-awkward time to introduce this as a best practice. It's not an admission of guilt, but it is an admission of suspicion.

Fortunately for this company, the EU has new rules requiring businesses to implement tools for exactly this kind of whistleblower reporting. It's awkward to be the first company to use a tool like this, but it's even more awkward to be the last. So this startup has a limited addressable market and a tricky go-to-market, but has the benefit that it won't take long at all to see whether or not it's going to work.

Supply/Demand Balance

It's getting slightly easier to buy or rent Nvidia GPUs ($, The Information). Importantly, this doesn't immediately affect Nvidia's economics: the two solutions to a shortage are higher prices or rationing, and Nvidia traded some immediate economic upside for more control over its future by a) rationing, and b) in particular, directing some chips to specialized AI cloud computing businesses that would reduce the biggest cloud companies' control over the market. So while this story is marginally negative for Nvidia, it's a testament to the company's management that this doesn't affect current revenue, just future revenue expectations. GPUs are cyclical, and even an extreme cycle is rarely enough to turn a business into a secular grower forever. A big component of the value of a cyclical company is how they handle upswings—if they maximize current-cycle profits, they're trading that off against their ability to control their destiny during the next cycle.