Eighteen months after U.S. federal judge Leonie Brinkema found that Google had illegally monopolised the markets for publisher ad servers and ad exchanges, we now know what life after that ruling will likely look like.

The Department of Justice pushed hard for divestiture, although many industry analysts expected the court to stop short of a forced sale. Two weeks ago, Brinkema did exactly that.

What the newly unsealed remedies opinion adds is the practical detail: six years of behavioural rules aimed at loosening the ties between Google’s AdX exchange and the publisher ad server inside Google Ad Manager, and giving rivals a better chance to compete.

The remedies do not dismantle Google’s vertically integrated ad tech stack. Instead they attempt to make the interfaces between its ad server, exchange, publisher header-bidding tools, and Google-controlled advertiser demand interoperable and non-discriminatory.

For many publishers, the biggest financial upside may still come from the class actions seeking damages for Google’s monopolistic conduct in adtech. But the operational remedies offer some potential upside, too.

That upside may prove modest. After all, many of the remedies broadly resemble those Google itself proposed as it fought off a break-up.

The key remedies:

• Unbundling Google’s ad exchange and its publisher ad server — legally and technically.

• Interoperability of “Prebid” with separate buy- and sell-side integrations. 

• Publisher data portability.

• Non-discriminatory Google Ads bidding with a block on “self-preferencing.”

• Publisher pricing control.

• Impact on publisher yield.

• Independent technical oversight.

The questions they raise for media:

• Can publishers leave Google’s ad server without losing effective access to Google demand?

• Does Prebid, the open-source technology publishers use to let multiple exchanges compete for an ad impression, now become a stronger counterweight to Google?

• Does transparency and access to publishers’ own auction data make switching technology providers easier?

• Will the demand for Google Ads compete on fair terms if used outside Google’s own stack?

• Could independent adtech companies ever gain market share?

• Do these changes improve publisher yield?

1. AdX demand becomes more portable

This is potentially the biggest change.

Google must allow its AdX exchange to submit live, real-time bids into competing publisher ad servers, breaking the technical tie between access to AdX demand and use of Google’s own DFP (its publisher ad server within Google Ad Manager).

The U.S. Department of Justice says this should give publishers access to important Google demand without forcing them to use both Google products. 

That reduces one of the biggest risks of leaving Google Ad Manager. Theoretically, at least.

The commercial test, however, is much tougher: Does a publisher moving to another ad server retain comparable access to Google demand — and comparable revenue?

Digiday’s survey of publisher and sell-side adtech executives rated this as one of the more significant remedies. But it’s accompanied by doubts about how this will work in practice.

2. Prebid gets a stronger role

Google must build separate integrations between AdX and DFP with Prebid, the open-source industry standard that publishers use to run competitive real-time bidding outside Google’s stack.

That could make Google demand compete through more neutral technology rather than receiving privileged treatment inside Google’s own system.

Again, it will come down to the details of implementation. A tiny increase in the milliseconds it takes systems to exchange bids materially affects who wins an auction.

3. Publishers get more transparency and control over their data

Google must allow publishers to access, audit, and export more of their data from both the AdX exchange and the DFP server. 

This includes being able to export their ad-server configuration and historical performance data, along with APIs and technical documentation to competing providers.

This matters because data portability should reduce the practical cost of switching providers.

For sophisticated publishers, this could make independent yield analysis, vendor comparisons, and testing alternative ad servers considerably easier.

4. Google demand remains the big attraction

Google Ads (formerly AdWords) is one of the world’s largest pools of advertiser demand.

The ruling says demand cannot favour AdX or other Google adtech products merely because Google owns them. And Google Ads will not be able to bid directly into DFP/GAM. 

But the court has not removed Google’s enormous underlying advantage in controlling advertiser demand. The crucial test is whether that demand now competes on genuinely equal terms whenever publishers use technology outside Google’s own stack.

That is the biggest reason the ad industry remains cautious. 

AdExchanger notes that Google may still have legitimate performance reasons for treating transactions differently (for example, around data, fraud prevention, privacy, or inventory quality). Determining where legitimate optimisation ends and self-preferencing begins will be difficult. 

One useful industry analogy is that the court may be making the pipes more equal, but Google still controls most of the water.

Google is also prohibited from restoring “first look” or “last look” auction advantages, which it used until 2019 to give its exchange privileged information or timing over rival bidders.

5. Don’t assume this immediately means more revenue or less cost

None of this automatically delivers higher advertising prices, lower fees, or greater publisher revenue.

In fact, separating AdX and DFP under separate contracts could increase costs as DFP fees were previously absorbed by exchange credits.

Publishers changing systems face migration costs and operational risk. In reality, switching tech on this scale is a mountain many media won’t be willing to die on. 

Some remedies concern practices Google has already altered. And the open-Web display market at the centre of this case itself has been deteriorating while the litigation dragged on.

So, you’d expect a lot of testing before major players choose to migrate, if they do.

6. Six years of technical policing

The remedies will be overseen for six years (the Department of Justice had asked for 15) by a court-appointed monitor and technical committee. 

That matters because adtech discrimination has proved to be extremely subtle. A competitor can theoretically have access while still being disadvantaged by slower responses, poorer data, or different auction mechanics.

Some of the most important compliance battles may therefore be about the technical interfaces through which software systems communicate — data fields and milliseconds — rather than dramatic courtroom arguments.

What comes next

The judge’s opinion is public, but her enforceable final judgment still has to be written.

Google and the Justice Department have until early October to submit their proposed wording for the final judgment. Brinkema will then settle the enforceable obligations and deadlines. 

Google, meanwhile, has flagged that it plans to appeal the underlying liability ruling concerning Google Ad Manager, according to Reuters. 

It will take time to see whether the remedies move the market.

Do major publishers begin to test alternative ad servers? Does Prebid win more business? Do independent exchanges gain share? Does AdX finally face pricing pressure? And can publishers move meaningful inventory outside Google’s publisher stack while retaining equivalent Google demand and equivalent yield?

If it’s a yes to any, these remedies could matter substantially.

If not, Google will have kept not only the structure of its adtech business, but much of the economic power that structure created.

Importantly, nothing in the likely final remedies will change the larger financial story: Google has already been found to have unlawfully monopolised publisher ad-server and ad-exchange markets. 

The major publisher class-action damages cases built around that conduct continue in many countries.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .