Scott Messer, a Los Angeles consultant who advises publishers on programmatic revenue, predicted that every one of the 19 conduct provisions in Google's September 2025 proposed order would appear in Judge Leonie Brinkema's remedies ruling. His scorecard, posted on LinkedIn after the 106-page opinion was unsealed on September 16, 2026, shows that all 19 did, three of them in modified form, while rating only four as high-value for publishers.

In Short

A publishing consultant compared the rules a judge has now placed on Google's ad business with the rules Google itself offered to accept, and found every one of Google's offers in the ruling. That matters to anyone who sells ads on a website, because these rules decide whether Google's advertisers can reach pages that do not run Google's own ad server. By his ratings, the changes that cost Google least come first, while the few that could move money toward rival ad tech are about a year away.

A scorecard built on Google's own filing

The post opens with a six-word verdict: "Clean Sweep with a few upgrades." Messer, principal and founder of Messer Media and co-founder of AdTech Therapy, a consultancy working on the sell side of digital advertising, added that the unsealed opinion held "nothing too surprising, but it was certainly interesting to read Judge Brinkema's rationale." Three images accompanied the text: his predictions with results, a table of additional remedies, and handwritten notes "from when I read all 106 pages."

A five-page export of the spreadsheet holds both of those tables, along with an earlier summary and a page of predictions. Its document title is "Google Remedy Sheet AdTech Therapy", and every page carries the heading "U.S. VS. Google AdTech Trial: Remedies". Each provision is graded on three scales - a shift type (Major, Minor or Easy Give), Google friction and publisher value - and given a deadline under a column headed "Not Later Than". A free-text column sets out what Messer calls the detailed ecosystem effect.

The provenance of each provision is traced through three filings in Case 1:23-cv-108 in the Eastern District of Virginia. Document 1431 is the initial remedies proposal Google filed on May 5, 2025. Document 1664-1 is the proposed order Google attached to a memorandum on the legal framework for remedies, filed on September 5, 2025. Document 1858 is the memorandum opinion, signed on September 2, 2026 and unsealed by a one-page order on September 16.

Two labelling problems need flagging. The first two pages are headed "Summary of DOJ-Proposed Remedies and Effects", yet the court docket lists Document 1664-1 as Google's proposed order, and the remedy text quoted in the sheet closely tracks the wording of Section III of that Google filing. The 19 provisions are Google's proposals, not the Justice Department's. Likewise, a column headed "In Sept 2025 Final Order?" refers to Google's proposed final order, not to anything the court entered. The post itself carries no date. LinkedIn displayed it as a week old on September 27, and the PDF of the sheet carries a creation timestamp of September 17, the day after the opinion became public.

Eight in May, 19 by September

The prediction page shows how far Google's offer grew between its two filings. Eight of the 19 provisions were in the May 2025 proposal, which PPC Land covered when both sides filed on May 5: real-time AdX bids for rival publisher ad servers, a direct request route for those servers, an end to restrictions on bid sharing, rescission of the policies behind them, a ban on conditioning AdX on DFP, the ban on First Look and Last Look, deprecation of Unified Pricing Rules, and per-exchange price floors.

Eleven more had appeared by September. They include a direct route for Prebid instances to request AdX bids; a rule against charging a higher AdX revenue share on traffic arriving from outside DFP; three data-portability items; bid data passed to a rival ad server at the publisher's request; written notices telling customers that the AdX and DFP bundle is optional; separate contracts for each product; Header Bidding Trafficking for every DFP customer regardless of size or Ad Manager 360 subscription; a server-to-server connection between DFP and Prebid Server; and a ban on DFP fee discounts tied to uniform floors. On the Prebid item, the sheet records that the May proposal "mentioned rival ad servers generally, not open-source Prebid."

Messer predicted "Accepted" for all 19. The court adopted 16 as written and three with modifications, according to his results page. Two of the three data-portability items were confined to open-web display after the court rejected the government's attempt to extend the export to video and in-app data. The DFP-Prebid Server connection went the other way: it was expanded to carry programmatic direct deals, meaning Programmatic Guaranteed and Preferred Deals, as well as indirect demand. Both changes match PPC Land's reading of the unsealed opinion, which recorded that the parties' proposals overlapped so heavily that most behavioral remedies were not seriously contested.

Where Messer sees little cost to Google

The ratings are where the sheet adds something the opinion does not. Of the 19 provisions, Messer classes eight as Majorshifts, six as Minor and five as Easy Give, his label for concessions that cost Google little. Friction for Google is rated Low on 10 and High on four. On publisher value, four provisions rate High, 11 Medium and four Low.

The four high-value items form a single cluster: the obligation to make AdX bids available to non-Google ad servers, the direct request route for those servers, the Prebid route, and the revenue-share rule. Messer calls the last of these a "Critical anti-discrimination rule", arguing that without it Google could render the rival-server and Prebid routes useless through penalty pricing or higher take rates on traffic that does not pass through its own ad server. All four also carry a High friction rating. Nothing else among the 19 does.

The Easy Give entries are blunter. On the written notices, the sheet states: "Sending email notices costs Google virtually nothing and produces zero immediate behavior change. Inertia and migration complexity keep publishers on GAM, not a lack of legal awareness." On the documentation that rival ad servers need to import DFP data, it is shorter still: "Writing technical documentation is cheap." In Messer's reading, Google "can fully comply with this provision while keeping proprietary schemas complex or cumbersome to map into rival ad servers in practice." The ban on fee discounts for uniform floors draws a similar objection: "Google can easily restructure overall base pricing tiers or service bundles to achieve similar commercial pressure, making this rule difficult to audit."

Items rated higher still come with caveats. Ad server migration, according to the sheet, "remains a complex, high-risk operational task, so data access alone won't trigger mass migrations unless alternative servers offer clear feature advantages." Separate contracts stop bundle terms in enterprise agreements, yet "Google sales teams can still cross-sell products aggressively through soft relationship incentives." Even the ban on tying carries a cost note for publishers: it "May increase prices for publishers if discount bundles cannot be offered." Extending Header Bidding Trafficking, a feature already built into Ad Manager, to smaller tiers "costs Google almost nothing while allowing them to claim broad support for open header bidding." On the server-to-server Prebid connection, Messer leaves open whether it strengthens Google's Open Bidding or makes it irrelevant.

The deadlines sharpen the picture. Under the dates in Google's proposed order, five obligations fall due within three months of the effective date: the three data-export items, the policy rescission and the written notices. Four of those five are Easy Give entries with Low friction. All four high-value provisions sit at 12 months in the Google filing, alongside the bid-data item and the DFP-Prebid Server connection.

Those dates are Google's proposals, and not all of them survived. The court set the effective date at 60 days after entry, not the 120 days in Google's order; left the deadline for AdX bids into rival ad servers to negotiation, since the government wanted six months against Google's 12; and expects the behavioral remedies to be in place within 15 months. Whether the new integrations meet the opinion's functional equivalence standard is a separate test that the scorecard does not attempt.

Pricing rules, rated low

Unified Pricing Rules, among the most cited practices in the liability case, rate Low for both friction and publisher value. The explanation sits in a note on the prediction page: "Already changed in EU as of Jan 2026". The companion provision restoring per-exchange floors rates Medium, and its entry holds the most concrete commercial statement in the sheet: "Publishers can now charge higher price floors on Google demand (to offset Google's fees) or offer lower floor prices to independent DSPs/exchanges." The plaintiffs' liability expert, Tim Simcoe, had calculated that eliminating the rules alone would cut AdX's fee from 20 percent to roughly 16.6 percent, a level the court compared with some competing exchanges.

A few days before September 27, Messer shared a post by Kean Graham that opened: "Google just handed publishers back a lever it took away in 2019." His comment: "That was fast! Granted, this feature was already built for the EU in January 2026, but it's nice to see some hustle." The lever in question dates to the 2019 shift in which Google retired existing floor rules on the move to a first-price auction and replaced them with unified pricing rules.

PPC Land's own earlier account does not line up neatly with that timeline. Reporting on the September 2 order, PPC Land wrote that Google removed Unified Pricing Rules from Ad Manager in December 2025 under antitrust pressure, restoring buyer-specific floors, and did not describe a regional limit. Messer places the change in the European Union in January 2026 and treats late-September availability as new. Neither the sheet nor the posts establish when per-bidder floors reached publishers in the United States, and no statement from Google on the rollout is among the source documents.

Six obligations Google did not offer

The fifth page lists what the court added. Google's proposed order was narrow by design. Its applicability section stated that nothing in it applied to AdSense, AdMob, Google Ads, DV360 or any Google tool other than Google Ad Manager, and that it covered only open-web display inventory. Three of the six additions reach past that line into AdWords, the advertiser product whose demand sat at the centre of the liability finding.

The first bars "any direct bidding integrations between AdWords and DFP through which AdWords could provide real time bids directly to DFP for indirect open-web display inventory", while leaving Google free to integrate AdWords with Prebid. Messer rates it a Major shift with Medium friction and value, and says it "Stops Google from using direct server-side paths (like gBid/Buyer Direct, Agentic) to channel Search advertiser spend exclusively into DFP, preserving exchange-level competition."

The second prohibits AdWords from "prioritizing bidding on AdX over other ad exchanges for indirect open-web display ad inventory solely based on the ownership of the ad exchange". It is one of only two additions rated High on both friction and publisher value. Messer's own entry qualifies it immediately: the rule gives rival exchanges access to AdWords liquidity, "However, AdWords can preference better signals/info/non-fraud inventory (ie, yup... AdX)." The opinion permits AdWords to prioritise AdX to achieve advertisers' objectives, and the factors that qualify are left for the parties to enumerate.

The third restricts Google from using first-party data, described in the sheet as Search, YouTube or Gmail user data, to personalise an open-web display impression "unless it does so without regard to whether the [real-time bid request] is submitted by AdX". Rated Minor, it "Codifies status quo while closing off future circumvention paths," according to Messer.

Rajeev Goel of PubMatic reached a comparable conclusion from the exchange side. In his assessment published on September 17, he singled out the AdWords provision as the most consequential in the order. Neither account alters the scope point that emerged when the opinion became public: DV360, Google's demand-side platform, was left outside the order entirely, the court having found an insufficient connection between it and the conduct at issue.

The remaining three additions concern transparency, oversight and reach. Google must supply DFP documentation and "a data file showing, for each ad served, the ad candidate prices and any price adjustments used in executing the DFP 'final logic'", due in 12 to 15 months with a status report in 90 days. Messer treats it as the replacement for the rejected open-source auction and gives a two-part verdict: "Most likely unusable, but increases trust via transparency."

The oversight structure - an independent Monitor, a three-member Technical Committee and an internal compliance officer at Google - is rated Admin/Easy. It "Creates an accessible forum for publishers and ad tech rivals to surface technical non-compliance or latency penalties directly to neutral experts." Google's own order had proposed a Monitoring Trustee whose provisions would expire on the third anniversary of the effective date. The court instead chose a Monitor who serves until the judgment expires, with power to inspect source code and interview Google staff.

Reach is the other High-High entry. The sheet quotes the opinion: "the Final Judgment will apply beyond the United States and reach globally... Google's proposal for a 6-year term is appropriate." Messer's effect note reads: "Ensures international publishers receive identical access to AdX/DFP integrations, preventing Google from maintaining regional moats outside the US." For European publishers that runs in parallel with Brussels, where Google rejected the Commission's breakup demand on November 13, 2025 and offered product changes instead. Counting the additions, the sheet covers 25 provisions, and six of them rate High for publishers.

The agentic question

Messer's post names one open issue. "I think the big thing to watch still is how the reality of AdX/Prebid/Adwords plays out, and if this hampers Google's agentic (open web display) media buying in any ways," he wrote. The only other reference to agents in the sheet is the parenthesis in the entry on direct AdWords-DFP integrations, which lists "Agentic" among the server-side paths the rule closes for indirect open-web display.

The court's record on the subject is thin. The opinion concluded that AI had not yet disrupted ad tech the way it had disrupted search, while noting that Google Partner Bidding, also called gBid Direct, already lets Google's buying tools bid directly into publisher auctions for mobile app inventory. Google's visible agent work has sat on the management layer. Ask Advisor, presented at Google Marketing Live on May 20, 2026, orchestrates specialist agents across Google Ads, Google Analytics, DV360 and Merchant Center.

Would an agent that places Google Ads budgets into open-web display count as AdWords bidding into DFP, or as something the order does not name? Neither the sheet nor the post answers that. The jointly proposed final judgment, due on October 2, is the first document in which such definitions could appear.

Why a publisher adviser backs the ruling

"In short, Google has to do all of the behavioral things and does not need to sell AdX or DFP," Messer wrote. He endorsed the outcome, arguing that Brinkema was right to reject the structural remedies because no likely buyer had emerged and a breakup "would be catastrophic (at least in the short term) to publishers." He also agreed with her treatment of the proposed escrow fund: "She argues why even an escrow fund to offset the costs wouldn't be legally or functionally sufficient too." The plaintiffs had asked for 50 percent of the net revenues of AdX and DFP, counted from April 17, 2025, to be placed in escrow, and the court found the lack of specifics fatal.

Other reactions have been more conditional. Digital Content Next framed the outcome as a test of whether Google would deliver what it had promised, and PubMatic, itself a plaintiff in a private suit, set four conditions for the Prebid route to count as meaningful choice. Messer's commercial position is on the publisher side. He has served since October 2024 as fractional head of identity and partnerships at The Arena Group, a publisher running a dedicated marketplace on Index Marketplaces that his profile promotes under the name Encore, and since August 2021 as executive in residence at Progress Partners, a Boston and New York investment bank working on mergers and acquisitions for emerging technology companies. He also co-hosts The Middlemen Podcast. His post closed with a line for both audiences: "Onward and upwards pubs. Vendors, start your engines."

PPC Land has cited Messer before, including his January 5 column in AdExchanger on publishers abandoning dependence on platforms, and his endorsement of IAB Tech Lab's bot management guidance in May.

Why the count matters for buyers and sellers

The scorecard is one consultant's grading, not an audit, and its friction and value ratings are judgments rather than measurements. Its usefulness lies in the ordering. For a publisher deciding whether to test a rival ad server, the provisions that change where Google's demand can flow are few, and on Google's own timetable they arrive last. For a media buyer, the AdWords rules are the ones with a measurable outcome.

The opinion itself supplies the baseline. Since the 2015 AWBid feature, AdWords has integrated with about 50 third-party exchanges yet placed only about 3 percent of its impressions outside AdX, and the share of AdWords open-web display transactions on non-Google exchanges fell from 11 percent in 2022 to 5 percent in 2024. If the ownership rule works as the court intends, those figures are where it would show. If Messer's parenthetical about signals and non-fraud inventory proves right, the numbers may barely move.

The calendar is fixed for the next step only. The parties must file one jointly proposed final judgment by October 2, 2026, with any disputed provision submitted in both versions and labelled by proponent. The judgment takes effect 60 days after entry, the Monitor and Technical Committee must be appointed, and the court expects the behavioral remedies within 15 months. Messer's sheet will need a new column when that filing arrives.

Timeline

  • 2019: Google moves Ad Manager to a first-price auction and retires existing floor rules in favour of unified pricing rules
  • January 24, 2023: The Justice Department and state attorneys general file the ad tech antitrust suit in the Eastern District of Virginia
  • April 17, 2025: Judge Brinkema finds Google monopolized the publisher ad server and ad exchange markets for open-web display
  • May 5, 2025: Both sides file initial remedy proposals; Google's is Document 1431, containing eight of the 19 provisions in Messer's sheet
  • September 5, 2025: Google files its remedies memorandum with a proposed order, Document 1664-1, containing all 19 provisions
  • September 22 to October 6, 2025: The remedies trial hears 26 witnesses over 11 days
  • November 3, 2025: The parties file post-trial briefs and competing proposed final judgments
  • November 13, 2025: Google rejects the European Commission's breakup demand and proposes product changes
  • November 21, 2025: Closing arguments conclude, with the bench questioning an AdX divestiture
  • December 2025: Google removes Unified Pricing Rules from Ad Manager, restoring buyer-specific floors
  • January 2026: Per-bidder floors built for the European Union, according to Messer
  • January 14, 2026: The European Commission publishes the provisional public version of its ad tech decision
  • May 20, 2026: Google presents Ask Advisor, an agent spanning Google Ads, Google Analytics, DV360 and Merchant Center
  • September 2, 2026: Judge Brinkema rejects all three structural remedies and files the memorandum opinion, Document 1858, under seal
  • September 16, 2026: The 106-page opinion is unsealed
  • September 17, 2026: PubMatic publishes its assessment, singling out the AdWords provision
  • September 17, 2026: Creation timestamp on the PDF of Messer's remedies sheet
  • Mid-September 2026: Messer posts "Clean Sweep with a few upgrades" on LinkedIn; the platform showed the post as one week old on September 27
  • Late September 2026: Messer shares Kean Graham's post on per-bidder floors in Google Ad Manager
  • October 2, 2026: Deadline for the parties to file one jointly proposed final judgment

Summary

Who: Scott Messer, principal and founder of Messer Media and co-founder of AdTech Therapy, assessing the remedies ordered by Judge Leonie M. Brinkema of the United States District Court for the Eastern District of Virginia against Google LLC, with publishers, independent exchanges, rival ad servers and advertisers using AdWords as the affected parties.

What: A five-page remedies sheet and LinkedIn post showing that all 19 conduct provisions in Google's September 2025 proposed order were adopted, 16 as written and three with modifications, alongside six provisions the court added. The sheet rates only four of the 19 as high-value for publishers and five as low-cost concessions, and mislabels Google's proposals as DOJ-proposed.

When: The post appeared in mid-September 2026, after the opinion was unsealed on September 16; the sheet's PDF is timestamped September 17. The jointly proposed final judgment is due on October 2, 2026.

Where: LinkedIn, concerning Case 1:23-cv-108 in the Alexandria Division of the Eastern District of Virginia, with obligations that apply globally for six years.

Why: The sheet sorts the remedies by what they cost Google and what they return to publishers, showing that the provisions most likely to redirect Google's demand toward rival ad servers and Prebid carry the longest deadlines, while several early obligations are, in Messer's assessment, paperwork.