Google updated two pages of its search documentation on 18 September, and the change is one line long. Local businesses now sit alongside hotels, flights, long distance trains or buses, and products in the list of queries that can trigger the aggregator unit and the supplier unit. One line, and it extends a brand new results architecture out of travel and retail and into every plumber, dentist, clinic and law firm in the European Economic Area. The Google supplier unit is the half of that architecture worth caring about, because it is the first search feature in years that hands the direct provider a block the intermediary cannot take from it.
Then you read the mechanics and find the catch, which is a dependency rather than a cost. Google's documentation is explicit that the supplier unit only appears if the aggregator unit appears. Your slot as the business that actually sells the thing is conditional on an approved comparison site or directory winning a block on the same query first. That inverts most of what practitioners have learned about intermediaries in search, and almost nobody has priced it in yet.
That fifth category is the reason this is worth an hour of attention rather than a bookmark. Hotels and flights have had comparison layers for twenty years, and the businesses in them have long since made peace with being listed by somebody else. Local services have not. A regional plumbing chain has never had to think about whether a directory is eligible for a search feature, because its visibility was its own problem to solve. As of 18 September, in the EEA, part of it is not.
What the Google supplier unit actually is
Two units, introduced to comply with the EU Digital Markets Act, sit together on the results page. The aggregator unit gives an approved Vertical Search Service a block of its own listings: think comparison shopping services, metasearch engines and directories. The supplier unit sits alongside it and shows the direct providers themselves, which Google's documentation defines as individual hotels or airlines, brick and mortar business owners, or providers of services such as plumbing.
The entry requirements are where the two diverge sharply, and the asymmetry is the story. An aggregator has to be approved as a Vertical Search Service, supply the necessary data, and meet quality standards to participate. For local business queries that means completing an interest form and providing a point of interest feed carrying name, address, phone, category, images and ratings. For flights and long distance transport it means direct feed integrations or real time APIs. A supplier, by contrast, provides nothing. Google crawls what is already there.
| WHAT DIFFERS | AGGREGATOR UNIT | SUPPLIER UNIT |
|---|---|---|
| Who it is for | Approved Vertical Search Services: directories, comparison shopping services, metasearch engines | Direct providers: an individual hotel, an airline, a shop, a local service business such as a plumber |
| Approval needed | Yes. Formal approval as a Vertical Search Service, plus stated quality standards | None documented. Eligibility follows from being the direct provider for the query |
| Data you must supply | A point of interest feed for local queries, or direct feed integrations and real time APIs for flights and long distance transport | Nothing beyond what is accessible through web crawling. Optional feeds can enrich the result but are not a condition of appearing |
| How many can show | One unit at a time, with the top ranked provider expanded by default and the rest behind a switcher | Shown alongside the aggregator unit when that unit renders |
| Can it appear alone | Yes. The aggregator unit is the trigger, not the passenger | No. The documentation states it only appears if the aggregator unit appears |
| Where it runs | EEA users only, for the five listed query categories | EEA users only, tied to the same query categories |
Look at the last two rows together, because that is the whole argument. The cheap unit is the dependent one. Google has built a feature that costs a direct provider nothing to qualify for and cannot be triggered by that provider under any circumstances. The only actor who can summon it is the intermediary whose business model depends on standing between you and the customer.
The dependency nobody has priced in
Work through what the dependency means in practice and it splits into several distinct outcomes, only one of which anybody is planning for. The comfortable case is a category with a strong approved aggregator: the unit fires, you appear beside it, and you have gained a block of owned real estate you did not have in August. The uncomfortable cases are everything else.
The second point deserves the most scrutiny, because it is the one that could go badly. Concentrating default visibility in whichever aggregator ranks first is efficient for the results page and risky for everyone listed inside it. If that provider's coverage of your city is thin, or its category taxonomy files you in the wrong bucket, the unit renders an incomplete picture of your market and your supplier slot inherits the framing. You cannot appeal a taxonomy you do not own.
Why the Google supplier unit costs less to enter than the aggregator unit
There is a reading of the zero cost entry that is too generous, and it is worth puncturing before somebody builds a quarter around it. The supplier unit asks for no feed because it does not need one. Google already holds the crawl of your site, whatever structured data you publish, and whatever entity record it has assembled about your business. Free entry is not generosity, it is confirmation that the data collection already happened and that the quality of what Google holds is now load bearing in a new place.
Which means the preparation work is unglamorous and entirely familiar. The crawlable facts about your locations, services, hours and prices are the input. If those are inconsistent across your site, your location pages and your markup, the supplier unit will render that inconsistency at the exact moment a customer is comparing you against an aggregator's tidy, feed driven listing. The asymmetry cuts both ways: the aggregator had to meet a quality standard to get in, and you did not, which means nobody checked your data before it went on display.
| WHAT GOOGLE WILL USE | WHERE IT COMES FROM | THE FAILURE YOU SHOULD LOOK FOR FIRST |
|---|---|---|
| Business name, address and phone | Crawled pages, markup, and your wider entity footprint | Suite numbers and trading names that differ between your location pages, your footer and your markup. Small mismatches, high visibility |
| Category and services offered | On page copy and structured data | Service pages that describe what you do in marketing language rather than in the category terms a searcher uses. The aggregator beside you will use the plain words |
| Opening hours and availability | Crawled pages and optional feeds | Hours maintained in one system and published from another. This is the single most common stale field on multi location sites |
| Images and ratings | Crawled assets, optional enrichment feeds | Aggregators supply these through an approved feed with stated quality standards. Yours arrive by crawl, so the gap in presentation quality is yours to close |
| Prices and offers | Crawled pages and product markup | Prices that live only inside a booking widget or a script. If it is not crawlable it does not exist for this purpose |
None of that is new advice. It is the same discipline that makes machine readable source data an infrastructure problem rather than a markup chore, applied to a surface that did not exist three weeks ago. The reason to revisit it now is that the audience changed. A stale field on a location page used to cost you a slightly worse listing. In a unit designed to sit beside a professionally maintained feed, it costs you the comparison.
What two different result architectures do to your reporting
There is a second order consequence that will land on analysts before it lands on anyone else. EEA results for these five categories now have a structurally different layout from results everywhere else. A query that returns an aggregator unit and a supplier unit in Dublin returns neither in Denver. Any report that averages position or click through rate across both is now blending two different pages and calling the result a trend.
That problem is not hypothetical, and it compounds a measurement environment that was already unstable. Third party positional data has had a rough few weeks for unrelated reasons, which we went through in detail when rank tracking quietly started dropping pages. Layering a region specific results architecture on top of that means a European position change in these categories can now mean at least three different things, and the export will not tell you which.
“A metric that mixes two different results pages is not a metric. It is an average of two answers to two different questions, presented with the confidence of one.”
The fix is boring and works: segment EEA traffic for the affected categories before you need to, not after somebody asks why Europe looks strange. Split the property in Search Console reporting by country, annotate 8 and 18 September, and keep the EEA series separate from everything else in these five verticals until the rollout settles. The regulatory divergence that makes this necessary is the same one we argued was worth watching closely in how the EU's search changes play out, and the reporting hygiene is the part most teams skip.
How to get ready before the deadline lands
Timing matters here. Search Engine Journal reports that the European Commission set a 60 day compliance window running from 23 July 2026, which puts the end of it within days of this documentation update. Expect movement rather than a quiet rollout, and expect the local category to be the noisiest part of it, because it is the one with the least settled aggregator layer.
See where you are cited today
A free snapshot audit of your rankings and AI citations before we ever talk.
Tyler leads work at the intersection of SEO and generative engines at Something Inc., helping B2B brands get ranked and cited across every major AI engine.