The Latest on the Google Spin-Out

The big news this week came from the Google Search antitrust case where Judge Mehta essentially said that a) Chrome spin-out is too hard; and b) stopping distribution deals to Apple et al is too costly; so therefore let’s pretend the search trial didn’t happen and move along. For more on this:

Anyway, that’s not what this newsletter is about. There’s been a bit of movement in the effort to give me enough new material for a Yield sequel, so I want to answer all the questions I keep getting asked in one place.

Sort of. This week the EU fined Google €2.9 billion in their version of the ad tech case, and kicked the can down the road on any structural remedies. Previously the EU had clearly wanted a break-up as a remedy, and provided this helpful graphic, which I’ve included in this newsletter before just because it makes me laugh:

There was previous reporting by Reuters that the EU was hesitant to order a structural remedy and this decision was being considered in light of trade negotiations with the Trump administration. Trump even responded in a Truth Social post:

The EU did not rule out a structural remedy. They just sort of passive-aggressively asked Google to suggest that themselves. “You know what you did wrong, now tell me,” more or less. Here is the key paragraph of the EU statement (emphasis in the original):

I’m sorry, but this is ridiculous. This paragraph basically claims the only viable solution is to break the company up, but we’ll let the company propose its own solution in the next 60 days, then think about it some more, then tell you what we really want to do here.

There’s a very simple translation for this: We want the U.S. to act first so we don’t piss off Trump.

We’re gearing up for the remedies phase of the DOJ trial starting Sept. 22 in Virginia. It’s also Rosh Hashana and I’m volunteering as our industry’s shabbos goy, doing all the illegal writing and typing so you can be with your family.

Other than the release of my book — which we can all admit was the biggest news of the summer — the other new development on this subject was yet another delay in the Texas antitrust case until sometime in 2026. This case has been going on forever, has moved jurisdictions thrice (by my counting), and even engendered an act of congress (!) to help move it along*. Yet, it is still in wait-and-see mode, hoping that the Virginia courts will give more guidance about remedies. It’s not just the EU that’s kicking the can!

As we gear up for the remedies trial, there has been a flurry of court filings by both sides laying out arguments, listing witnesses and exhibits, and generally showing why lawyers bill by the hour.

The DOJ’s proposed remedies can be read in this 60-page filing. This is largely a refinement of previous filings from the spring and can be summarized as:

Exciting stuff! We’ll have more to say as the trial gets under way. In the meantime, let’s talk about the first bullet: AdX divestment.

As you might know, I’m technically a bidder for AdX through my GoFundMe. (I’ve got $271 pledged. Every donation helps!) But putting aside my conflict of interest, let’s talk about what’s for sale and speculate a bit about what it might be worth.

There has been a lot of chatter about how big AdX is, but we actually have a data point from the trial. A 2020 document was disclosed in the trial that shows GAM/AdX’s media revenue as $9.3 billion revenue and its ad serving fee revenue as $260 million. The chart below shows a bunch of internal accounting that I don’t fully understand (there’s a reduction of $1.47 billion for “PG/PD adjustment”?), but the important part is the net revenue in 2020 was $1.55 billion — $1.29 billion, if you remove ad serving fees.

We have to make some adjustments. First, this data is five years old. We know that the Network segment as a whole has barely grown in these years, and we also heard during the trial that the in-app AdMob side of things was growing the fastest. So in a highly optimistic scenario, this number has stayed the same. In all likelihood, it is lower than $1.29 billion.

Wait, did I say something about AdMob? While AdMob is a separate column in the financials, that doesn’t mean that the GAM line does not include any mobile or video ads. The court only found a monopoly in the market for “open web display ads,” not in CTV or anywhere else. Google could make a strong argument that they can continue to offer an exchange in anything that is not “open web display ads,” which will further cut back on the size of whatever is being sold.

So let’s say the net revenue of the AdX that’s actually for sale is $1 billion.

I always like to give my disclaimer that this is not a financial newsletter. However, we have two clear comps in this case: $MGNI ( ▲ 2.01% ) and $PUBM ( ▲ 2.15% ). In the 2020 financial docs, the operating profit of GAM/AdX was listed as $602 million, but I find it highly unlikely that number will hold up in a spin-out due to all kinds of internal accounting and removal of ad serving. Let’s just do revenue multiples since we have no idea how profitable the spun-out AdX will be.

There’s a big difference in multiple between these two companies, probably driven by Magnite’s story about being a CTV leader. However, as we said above, AdX might not have any video at all depending on how Google carves up the assets to be spun. So let’s say 3x revenue for AdX, which would imply enterprise value in the range of $3 to $4 billion.

This valuation is probably too high, though, given growth trends. The spun-out AdX is “open web” only, which is in decline. And, crucially, the vast majority of demand coming into AdX comes from other Google sources, namely Google Ads and DV360**. Post spin, you would expect those demand sources to reduce significantly as they bid into other exchanges, seeking the best ROAS without regard to inter-company subsidies.

The net result here is that a buyer of AdX might be grasping at a falling knife. A shrinking business that is poorly positioned versus direct competitors, with a former owner that represents more than 50% of revenue, and whose top product advantage is being forcibly removed by the courts.

This implies that the best valuation for the spin would not be a financial buyer, but rather a strategic with a ton of ads demand to push through those same pipes in the way Google has for a decade. But how would the court — and Google’s attorneys — look upon Microsoft, Meta, or Yahoo as a potential buyer?

* In 2022, as part of the omnibus spending bill, Congress passed the State Antitrust Enforcement Venue Act, which made antitrust actions of state attorneys general exempt from the standard venue consolidation process that other suits are subject to. This allowed Ken Paxton to move the case back to Texas from New York.

** Trial documents indicated that about 70% of DV360 demand goes to AdX, while 60% of all AdX demand comes from Google Ads. Google Ads currently only bids on non-AdX exchanges for retargeting campaigns.