What if?

A good day for some speculation. What if Google actually got broken up? Also great interviews with the CEO of Admiral’s (just raised $19 million) and the head of media at Roku.

The Vendor Interview: Admiral

Marketecture Vendor interviews are free for a week, then require a subscription. Read more about subscription options at marketecture.tv.

Podcast: Jay Askinasi on Roku’s past and future

What if?

Last week Bloomberg and others reported that the government was considering an aggressive remedy in the search antitrust case that could include a forced divestiture of Chrome, Android, or even Google Ads (AdWords). While this newsletter isn’t an antitrust deep-dive (this one is), the subject is too juicy for me to ignore.

Since Judge Mehta published his findings (our coverage is here) that Google has a monopoly in search and in search text ads the conventional wisdom has been that a) the likely remedies were going to focus on distribution of search into browser defaults; and b) that like the Microsoft case decades ago the remedies would be too little and too late to make a difference. Perhaps Google itself was pushing this train of thought.

The Bloomberg leak…ahem…story…suggests a more muscular effort on behalf of the government to both make a dent in the monopolistic behavior, and to bring a reluctant Google to the bargaining table.

Apple (and Mozilla) deals

The allegations in the complaint and the subsequent findings largely boil down to the virtuous cycle whereby Google gets more search data, and therefore can improve its search product, and therefore can pay for more distribution to get more data. A lot of attention then has gone to the $20 billion+ Apple distribution deal, which certainly is a big part of this cycle.

Undoing or regulating the Apple deal is a pretty interesting game theory problem, because the current equilibrium (Google gets searches, Apple gets money) is very stable and leaves both parties at their best outcomes. There are three dimensions to a remedy here:

1. Who can pay for default placement;

2. Whether there is a revenue share;

3. How does the consumer get to choose a search engine (inherently related to default choice).

Option 1: Google cannot pay for default distribution, and other parties are allowed to, but rev shares are permitted:

• Apple gets paid a lot less since the other search engines aren’t nearly as monetarily efficient;

• Consumers get a default search engine they don’t prefer;

• Google gets less traffic, but still probably quite a bit, and now for less cash.

Option 2: No one is allowed to pay for default distribution, consumers get a choice screen of some kind, but there is still a rev share:

• Unless prohibited from doing so, Apple would still be incentivized to promote Google since its the most monetarily efficient;

• Google saves money, maybe loses some share.

Option 3: Google is prohibited from paying anything for any kind of distribution (no rev share), other parties are allowed to monetize:

• Apple loses big as they can only get monetization from less preferred partners;

• In this scenario Apple is highly likely to build their own search engine;

• Google loses the most.

Ari’s opinion: Option 3 is the only one that makes a meaningful difference in the monopoly, but it’s also an option that has unknown and chaotic effects.

Chrome and Android spin-outs

According to the evidence in the DOJ trial, the Apple deal represents ~14% of Google searches, while 20% are from user downloaded Chrome and Android OEM agreements make up about 9.5%. So just solving the Apple distribution problem is not be enough. What to do?

The framework above doesn’t really work, since Google doesn’t pay for default position on these platforms. You could offer consumers a choice screen, but that’s a pretty unconvincing remedy. So maybe we should break it all up?!

Spinning out Chrome from Google has a fatal flaw — there does not appear to be a viable business model for a browser other than search revenue. Mozilla, the largest independent browser gets more than 80% of its revenue from Google search. If an independent Chrome was prohibited from selling default search position to Google, it would be hobbled from the start and all the issues relevant to Apple (above) would apply.

Chromebooks are estimated to bring in ~$5-7 billion in revenue, but presumably the majority of those numbers go to the OEMs. Maybe Google can get $1 billion in licensing fees, but that seems like a stretch.

Android, on the other hand, generates sizable revenue from its Google Play store, embedded Google services, and, of course, ad revenue (this article details all the ways Android makes money). What makes an Android spin really complicated is that all these services and revenue streams are interconnected. Even though Android is open source, OEMs install all the Google services on the devices to make them competitive for consumers with Apple’s complete, integrated offering. An independent Android (and the regulators forcing a spin) would be confronted with the question of whether to break Google’s hold on the Play Store, music, video, maps, the mobile ad network, and other apps. Given the narrowness of the judge’s decision (monopoly in search), this seems like significant overreach.

Ari’s opinion: Chrome spin-out is easy but economically doesn’t make a lot of sense, Android spin-out is phenomenally difficult. I don’t think this happens.

What about the text ads?

Judge Mehta found that Google has a monopoly in text ads, but not in all types of search ads. How would you solve the problem of text ads?

Option 1: Regulate the auction for text ads. This would require Google to submit all of it’s ad-related changes to government scrutiny. I think we all agree this would be a bad way of moving forward.

Option 2: Allow demand sources other than Google Ads to bid on searches. In this scenario Bing or other ad platforms would bid into searches with their own text ads and pricing, and presumably revenue share with Google when they won.

• Theoretically this could make the market for these ads more fair, since competitors could take lower margins to win more, or bring some targeting innovation;

• Google wouldn’t really lose out that much since they would still be getting paid;

• But would anyone other than Google Ads really win a significant share of auctions?

• And the technical and operational issues are daunting, to say the least.

Option 3: Spin out Google Ads

Hoo boy. Presumably you would only spin out Google Ads in conjunction with the text ads interoperability of Option 2, above, or else you would just have one bidder and it would still be a monopoly.

• Google Ads has a lot of stuff in it that has nothing to do with text ads. It is the only way to buy YouTube ads, for example. Separating this out would not be feasible so we would get YouTube interop as a side bonus.

• If a separate Ads was bidding into search, Google would still take a significant revenue share and capture the lion’s share of revenue. With interoperability, the independent Ads would have competitive pressure from Bing et al and you would expect margins to decrease. History has shown that independent bidders on search have been bad businesses (Kenshoo, Marin), so why would this be different?

• Ads performance would decline as it currently has access to enormous amounts of Google resources and data and as a separate company would have much less visibility.

Ari’s opinion: Option 3, an Ads spin out, could actually make sense. But it would be a seismic change and Google would fight to the death to prevent it.

Reading list

• Adelaide raises $1.4 million (link)

• Admiral raises $19 million to be a visitor relationship management (“VRM”) system – think CRM but for publishers (link)

• Netflix ties up with Infosum, Snowflake, LiveRamp to support clean rooms (link)

• Nexxen providing TTD with ACR data (link). Note: no one seems to know what this deal means, if you have any insight, reach out.

• Chick-fil-A launching a streaming service (link). Presumably will not be streaming on Sundays.

• Conde Nast licenses to OpenAI (link)