The Google/Apple search deal is always fun to read about, because the basic situation between them is:

1. Apple controls the default search engine settings on one of the two big global smartphone platforms, and it's the one that richer people use.

2. Google runs what is, in purely financial terms, the best search engine in the world; nobody can convert default search engine status into revenue more effectively.

3. The two companies are also direct competitors in mobile, and indirect competitors on many other axes. For example, Apple likes a world where apps are whitelisted and monetized by some central platform; Google prefers a wide open Internet whose fragmentation makes their search business more valuable. There was also a time in which many sophisticated investors believed that the vertical specific search engines that Apple was enabling through their apps platform would obviate the need for a general search interface—you wouldn’t Google “restaurants near me” when you could just use Yelp.

Every few years, these two companies have a meeting whose theoretical premise is to discuss whether or not Google should be the default search provider in iOS. In practice, Google is the default, and no one can afford to come close—Bing literally offered over 100% revenue share, but Apple turned them down because of quality concerns. Instead, they stuck with Google, which pays them 36% of search revenue instead.

These deals are always interesting because both parties have the maximum possible incentive to come to an agreement—Google+Apple is more valuable than Google-plus-any-other-handset-company, and more valuable than Apple-plus-any-search-engine—but that means either side could theoretically justify giving the other side almost all of the economics. They're playing the ultimatum game, a psychology experiment in which two players are offered a sum of money, and one of them can decide on how to divide it, with the other either accepting their division or rejecting it (in which case neither one gets the money). If two perfectly rational agents are playing the game exactly one time for exactly $1, the optimal strategy is for the one who divides up the money to offer $ε, and keep $1-$ε, and it's in other agent's interest to take it. Nice, a free $ε!

Human players don't do that, because we're wired for a world of iterated games. The real world doesn't offer clean examples of a one-shot ultimatum game, and instead gives us opportunities where the payoff is uncertain, there's always a potential future round, and where there's also an expectation on the part of consumers that both sides will come to a reasonable agreement. The people who have strong opinions about search engines will just fiddle with their defaults, but everyone else will notice that their search experience has been slightly degraded. (If you're voluntarily reading this article, and you've gotten this far, you are way, way more interested in search than the average person.)

The dynamic where consumers have an implicit veto over aggressive attempts to flip the economics towards one side is especially salient in carriage negotiations, when a cable network (e.g. ESPN) gets withheld from cable subscribers or more recently from streamers. Part of the TV bundle is the assumption that all the stuff you'd expect to have is, in fact, going to be available. One of the economic drivers of these deals is that from the customer perspective, Charter or Comcast or whoever is taking away their sports viewing, and cable companies are just not as sympathetic as media brands when it comes to disputes like this.¹

This is especially fraught in media because distributors and creators are so different. The company that owns distribution has shareholders who will have questions about any hit to margins, and they have plenty of lawyers and accountants who can elegantly structure deals. But they don't have rabid fans! So there's sometimes a sawtooth pattern to this: over time, the business side tries to get and exploit more leverage, and occasionally this blows up when the talent informs their legions of fans that some guy in a suit is making it harder for them to watch or listen to what they want.

That's part of what makes the Apple/Google situation so special: these companies are both popular. There's considerable overlap in their users, and even more if those users are weighted by revenue, and plausibly even more than that if they're weighted by the net present value of that revenue (people who pay up for fancier devices tend to use them a lot, and quality software and hardware are complements.)

Over time, these kinds of situations get more common, because a more complex economy has more narrow-but-indispensable niches. When some of these niches have compounding returns to scale, they tend to get monopolized, and when a niche shows the promise of fitting that pattern, it's increasingly easy for the company targeting it to raise whatever money it needs to race after that opportunity.

So, over time much more economic activity depends on a fairly small set of companies who have a lock on one specific stage in a process. That's a more profitable economy, and for those companies to maintain their position, they'll mostly tend to reinvest in making the product better. But it's also a more brittle one, where elaborate supply chains sometimes snap at a single critical link because someone, or both someones involved in the negotiation, get too greedy.

In The Diff, we've covered these negotiations across industries and scales:

• We've written more about the Apple/Google relationship ($).

• Search is prone to this because it's such an incredible business.

• One way to get an advantage is to raise money knowing you'll waste a lot of it but get priceless knowledge you can use to get better unit economics later.

• And sometimes great companies start as arbitrages but grow into something durable.

• This doesn't just happen between companies: it also explains why AI researchers get paid so well lately ($).

Join the discussion!

1  That's one reason media companies have gotten vertically integrated, unless they're a specific fragmented endpoint with a regulatory legacy, like movie theaters. If Netflix owns the Stranger Things franchise, they can just internalize the economic benefit of the complementarity between distribution and content. But if they have to shop for it, the Duffer Brothers know that they can ask for a high price.