Google Handed EEA Search to Aggregators. Direct Brands Get a Box With No Prices.
Google rolled out two new result units across the European Economic Area on September 8, 2026: an aggregator unit for approved comparison sites and a supplier unit for direct providers, covering hotel, flight, long-distance train and bus, and product queries. The supplier unit only appears when an aggregator unit appears, so direct visibility in EEA search runs through a comparison site. Google's 60-day compliance deadline lands around September 21.
Google's official line, delivered to Reuters and relayed by Search Engine Roundtable, is that this is "the largest reduction in quality of service" in Search's 29-year history. Nick Fox, who runs knowledge and information at Google, went further and said the changes "degrade the user experience for Europeans." Companies do not usually publish sentences like that about their own product. This one is aimed at Brussels, not at you.
The practitioner story is narrower and, I think, more useful. Read the two documentation pages Google published on the same day and a very specific hierarchy falls out.
Who gets the top of the page now
The aggregator unit documentation lays out the rules. Only approved Vertical Search Services get in: online travel agencies, comparison shopping services, metasearch engines, directories. They have to push data through Google's feeds or real-time APIs (a Lodging Point of Interest feed for hotels, the Transport features API for ground transport, the Partner Live API for flights, and a CSS integration for products). One aggregator unit per search. The top-ranked provider is expanded by default with photos, prices, and ratings. Everyone else is a collapsed tab the user has to click to swap in.
Then there is the supplier unit, which is where individual hotels, airlines, and product sellers go. Two lines in that doc matter more than everything else combined. First: "The supplier unit only appears if the aggregator unit appears." Second: you need no data beyond what Google can crawl to be in it, though results "can be enhanced with data feeds, if available."
So the shape of an EEA hotel query from this week onward is: one expanded aggregator, a couple of collapsed aggregators, and then a row of direct suppliers. According to Hospitality Today's breakdown, that supplier row now shows the hotel name, a link, the address, and a phone number. No live rate. No date filter. No "boutique" or "budget" tags. Everything except the one thing a traveler comparing hotels actually wants.
The closest analogy I can find is a shopping mall where the anchor tenant gets the main entrance and the independents get listed on a directory board near the restrooms. The board is accurate. It is also not where anyone makes a purchase decision.
The 30% number and the 0.8% number
Google's argument for why this is bad rests on one figure. In a September 2025 blog post, Oliver Bethell, Google's senior director for competition, wrote that "key parts of the European tourism industry have already seen free, direct booking traffic from Google Search plummet by up to 30%" after the first round of DMA changes. Google's own test markets (Germany, Belgium, Estonia) reportedly showed hotels losing more than 10% of traffic. The same post cites a study estimating up to 114 billion euros in revenue losses for European businesses.
The intermediaries have a different number. EU Travel Tech, the Brussels lobby for the travel platforms that benefit from all of this, points to a follow-up analysis from Mirai, the hotel direct-booking vendor whose data Google originally leaned on. Mirai's later figure put the overall impact at a 0.8% dip in direct bookings across Europe. Direct bookings through hotel websites only, and nobody has cleanly measured whether OTA bookings rose to offset it.
Both sides are lobbying, obviously. But the gap between those two numbers tells you something about metric types. Google is measuring free clicks from Search to hotel sites. Mirai is measuring completed bookings. A hotel can lose 30% of Google referral clicks and lose almost no bookings if those clicks were mostly people who were going to book through Booking.com anyway. From what I can tell, that is roughly what happened last time, and I would expect this round to look similar: a visible traffic drop in Search Console and a much smaller dent in actual revenue. Which is a lousy consolation if traffic is your KPI, and honestly, for a lot of in-house teams it still is.
Why Google is doing this now, on this timeline
On July 23 the Commission fined Google 460 million euros for self-preferencing its own shopping, hotel, transport, and sports results with richer visuals and filters than third-party rivals got, plus another 430 million on the Play Store steering case. Per TechTimes' report on the decision, Google got 60 days to comply or face periodic penalties of up to 5% of daily worldwide turnover. Sixty days from July 23 lands around September 21. Hence the September 8 rollout, and hence the documentation dropping the same day.
The regional differences page is the tell for where this goes next. It already maps a parallel set of features for South Africa and Turkiye: badges, refinement chips, structured-data carousels. Google is building a template for "regulator-compliant search" that it can drop into any jurisdiction that asks. And to be fair, this is not entirely new. Shopping in the EU has run through CSS partners since 2017. It just feels a lot less optional now that it covers hotels and flights too.
The uncomfortable move: get inside the aggregators
I keep seeing hoteliers and airline marketers treat this as a "protect direct" problem. I think that reading is backwards for at least the next quarter. The aggregator unit is organic real estate, and the only route into the expanded slot is through a Vertical Search Service that Google has approved. If you sell hotel rooms, flights, or physical products to EEA customers, the question stops being "how do I rank" and becomes "am I in the feed of the aggregator that ranks."
Here is the audit I would run this week, in this order.
1. Pull the exposure from Search Console. Filter Country to your main EEA markets, then filter Query to your commercial terms ("hotel," "flights to," product category names). Compare clicks for September 8 onward against the two weeks before. My working benchmark: if EEA clicks on those query groups are down more than 15% by September 22, with impressions roughly flat, the unit is doing it and no amount of on-page work will get it back.
2. Check which aggregators are firing for your queries. Search your top 20 EEA commercial terms from a European IP or a VPN and note which VSS is expanded by default. In most hotel markets it is going to be one of Booking, Expedia, or Trivago. That is your distribution list now, whether or not you like their commission.
3. Hotels: fix the rate feed before September 30. Hospitality Today reports Google stops sourcing hotel rates from third parties on that date. After that, rates come only from your own Hotel Center account or a connectivity partner, and even then they show up on Maps and google.com/hotels rather than in the Search unit. If you do not have a feed, your property has no price anywhere on Google in the EEA. That is a two-week job with a connectivity partner and it is not optional.
4. Products: run a CSS. This is the contrarian part for US ecommerce brands. Comparison shopping services are already the plumbing for EU Shopping, and the products version of the aggregator unit runs through them. A CSS partnership usually costs a monthly fee rather than a percentage, and it gets you into the unit's product pool. If you have ever looked at EU expansion and passed because of the shopping complexity, the calculus just moved. We covered how Shopping click-through is rising as AI Overviews absorb low-intent queries, and the EEA unit compounds that: fewer, better-qualified product clicks flowing through fewer doors.
5. If you are a marketplace or directory: apply. Google's form for ground transport, flights, and hotels is linked from the aggregator unit doc, and products go through the CSS contact form. Approval is not automatic and the data requirements are real. But there is a window right now where the list of approved VSS for niche verticals (long-distance bus, for instance) is short, and the top-ranked slot is expanded by default.
The supplier unit still matters, just less than you would hope. It is crawl-based, so the basics count: a clean entity on the site, consistent name and address, a crawlable page for each property or route, and structured data where Google's feed docs allow enhancement. That gets you the box. It does not get you the price.
What I'd bet on for Q1
My prediction: by the end of Q1 2027, the top-ranked aggregator in each EEA hotel unit will take more than 60% of that unit's clicks. It is the only provider expanded by default, and default-expanded wins on every interface I have ever seen data for. That concentration is going to make the ranking inside the unit the most valuable and least transparent piece of European SEO. Google's docs do not explain how the top slot is chosen. Nobody outside Google will know for months.
One expanded box, chosen by an algorithm nobody outside Google can see, now decides who gets the European hotel click.
I do not think the direct-booking apocalypse Google is describing shows up in revenue. Last time the click loss was real and the booking loss was tiny. But the teams that sat out the CSS and OTA feed work in 2024 because "we're a direct brand" are the ones most exposed here, and there is no version of this where a nicer landing page fixes it. Get into the feeds, measure what actually drops, and then argue about whether the DMA was a good idea.
Notice Me Senpai Editorial
