Skip to content
Research brief · Google in EuropeNovember 2026

Google Search impressions for European business profiles fell 83% after Sept 22

Direction requests barely moved.  The loss shows up in website clicks, about 9 percentage points worse than the same weeks last year, after Google's redesign of European results under the Digital Markets Act put directories and business websites at the top of many local searches.

Based on 273,000 locations in the European Economic Area and 1.2 million in the UK, Switzerland and the US, with daily data through Nov 9, 2026 and monthly history back to January 2024.

−83%
Change in Google Search impressions for EEA business profiles, mid-September to early November
98%
Share of EEA locations that lost more than half their Search impressions
−9 pts
Percentage points by which website clicks from EEA profiles trailed last year's change over the same weeks; direction requests stayed within a point
−10%
Change in Google Search impressions for the same brands' own location pages over the same weeks
Summary

Google Search impressions for business profiles in the European Economic Area fell 83 percent after Sept 22, when Google's new European layout reached local searches.  Customer actions fell far less than impressions, but website clicks took a real hit in some industries.

  • Every EEA country and industry lost Search impressions on the same days, while Search in the UK and the US, and Bing and Tripadvisor in the EEA, held steady.
  • On category searches, a block of directories often leads the page and Google's profiles sit in a collapsed row.
  • On name searches, Google counts far fewer impressions even though people keep searching.
  • Direction requests held in most industries.  Website clicks from profiles ran about 9 points worse than last year, and more than 20 points in insurance, finance and legal.

What to do.  Report EEA performance on calls, directions and website visits against the same weeks last year, watching website clicks most closely, mark Sept 22 on any chart of Search impressions, and keep location pages current, since Google's new layout links to business websites directly.

The three measures in this brief

Google reports how often a profile was seen in Search and in Maps, and how often someone acted on it with a call, a direction request or a website click.  Each comparison uses the same locations in both periods, Sept 7 to 20 against Oct 26 to Nov 8, with the same weeks of 2025 and the UK, Switzerland and the US as baselines.  Where we compare with last year, we give the gap in percentage points: website clicks that fell 10 percent this year against 1 percent last year are 9 points worse.  The EEA is the 27 EU countries plus Norway, Iceland and Liechtenstein.

Search impressions across the EEA fell to a sixth of normal within five weeks

If the drop came from particular accounts or listings, it would show up unevenly, hitting some countries, industries or accounts much harder than others and arriving at different times.  So we looked at when and where Search impressions fell.  They fell across the whole EEA on the same days, and held steady outside it.

Search impressions were normal through Monday, Sept 21.  By Thursday they were at half their usual level, and by the following Monday at about 30 percent.  They slid further through October and have stayed near 17 percent of the September level since late in the month.  All EEA countries moved on the same days, by 72 to 86 percent, while the UK and US stayed within a few points of where they started.

Phones and desktops dropped by similar amounts, 84 and 77 percent.  Maps impressions held for three weeks and then slid in late October, by about a quarter.  Maps also fell in the UK and US this autumn, by 8 and 11 percent, so part of the slide is not specific to Europe.  The rest may partly reflect fewer searchers opening Google's map from the Search page when a directory leads the new block, though we can't separate that from the season.

Among EEA locations with at least 100 Search impressions in the two weeks before, 98 percent lost more than half, and only 1 to 3 percent fell to zero.

Google Search impressions by week, indexed to Sept 1 to 21Fig 1
0255075100Sep 1Sep 8Sep 15Sep 22Sep 29Oct 6Oct 13Oct 20Oct 27Nov 3Sept 22UK 102US 101Switzerland 87EEA 17

Each market's Sept 1 to 21 average is set to 100.  Weeks run Tuesday to Monday, so the first week after the change starts on Sept 22.  The EEA was at 48 the week of Sept 22 and 17 by late October.  Switzerland, outside the EEA, settled near 87.

The EEA's share of Search impressions held steady for 32 months, then fell to 8%

If the drop were part of a pattern we had seen before, such as a seasonal swing or the March 2024 round of Digital Markets Act changes, something like it would appear earlier in the data.  So we looked back to January 2024, comparing EEA locations with UK locations to cancel out changes that hit both.  Nothing comparable appears in almost three years.

From January 2024 through August 2026, an EEA location drew between 36 and 54 percent of the Search impressions a UK location did.  In October 2026 the figure was 8 percent.  Google's first round of Digital Markets Act changes, in March 2024, did not move it.  In July 2025, Search impressions fell about a third in every market at once and recovered the following month, and the EEA's share was unaffected.

EEA Search impressions per location as a share of the UK's, monthlyFig 2
0%20%40%60%Jan 24JulJan 25JulJan 26JulOctMar 2024: first DMA changesJul 2025: global dipEEA 8%

Using a ratio cancels out changes that hit both markets at once, such as the July 2025 dip.

Swiss businesses near the EEA border lost up to a third of their Search impressions

If the change followed where the business is, Swiss businesses would be untouched, since Switzerland is outside the EEA.  If it followed where the searcher is, as a change made under the Digital Markets Act should, Swiss businesses that draw searches from across the border would lose impressions too.  So we looked at Swiss cantons.  Those around Geneva, Basel and Ticino, where many searches come from across the border, lost the most.

Switzerland is outside the EEA, yet its Search impressions dropped about 13 percent in the same week.  Broken out by canton, the drop is concentrated along the border.  Ticino, next to Italy, fell 33 percent and Geneva, surrounded by France, 30 percent.  Bern fell 6 percent and Glarus was unchanged.

Border cantons are where searches from people in Italy, France and Germany would be most common, so the pattern indicates the change applies to searchers located in the EEA, wherever the business is.  Businesses outside the EU that rely on visitors from it, such as those in border towns or tourist areas, can expect part of this drop as well.

Change in Google Search impressions by Swiss canton, Sept 7–20 to Oct 26–Nov 8Fig 3
Main cross-border regionsOther cantons
Ticinoborders Italy
−33%
Genevaborders France
−30%
Basel-Stadtborders Germany, France
−22%
Graubündenborders Italy, Austria
−20%
Schaffhausenborders Germany
−19%
Juraborders France
−18%
Zurich
−11%
Lucerne
−11%
Bern
−6%
Fribourg
−6%
Solothurn
−5%
Glarus
0%

Cantons with at least 300 locations.  Blue marks the cantons with the largest cross-border catchments.  Zurich and Solothurn also touch the border, but their main towns sit farther from it.  The grouping is our judgment.

Bing and Tripadvisor showed no drop on Sept 22

If European businesses were truly being found less everywhere, because of something in their listings or in how those listings are distributed, we would expect other publishers to show a similar drop.  So we looked at the visibility that Bing, Tripadvisor and EEA directories report for the same locations and weeks.  None of them dropped on Sept 22, so the change is specific to Google.

Bing listing impressions in the EEA rose 13 percent and Tripadvisor's were flat.  Apple reports engagement rather than impressions, so it is not in the table; its place card opens eased over the autumn with no break on Sept 22.  The EEA directories began falling the week of Sept 14, before Google added local searches to its new block, and gradually, unlike Google's one-week drop.

Change in visibility reported by each publisher, Sept 7–20 to Oct 26–Nov 8Table 1
PublisherWhat it countsEEAUKUS
Google Business ProfileSearch impressions−83%+3%−2%
Google Business ProfileMaps impressions−25%−8%−11%
BingListing impressions+13%−12%−5%
TripadvisorListing impressions−2%+15%+7%
EEA directoriesListing impressions−31%––

Same locations in both periods.  Google Maps impressions fell in all three markets this autumn, after rising in the same weeks of 2025, and fell most in the EEA.  Bing is compared one week later, Oct 19 to Nov 1, because its latest week had not loaded.  Directories are national and regional EEA listing sites, reported for EEA locations only.

Search impressions fell 70% in a week, direction requests held and website clicks slipped

If businesses were truly being shown 80 percent less, we would expect people to engage with them less at something like the same rate.  So we looked at calls, direction requests and website clicks, one at a time.  In the week of Sept 28, EEA Search impressions were 70 percent below their mid-September level and Maps impressions were unchanged, so total profile impressions were down 30 percent.  Direction requests were up 1 percent and calls down 1 percent.  Website clicks were down 5 percent, the one action that moved from the start.  Together, actions were down 1 percent, far short of the 30 percent fall we would expect if the lost impressions had been worth as much as the average impression.

Most impressions never had much chance of becoming an action.  A local search logs an impression for every business shown, while the searcher picks one at most, and someone who wants to call or visit a business still finds a way there, through Google Maps, a search for its name or the other results on the page.  Website clicks are the exception.  In Google's new layout, described below, a searcher can reach a business's website from a carousel of websites without passing through the profile, or land on a directory's site instead.  Clicks on the same brands' websites have not risen so far, so we can't yet say which matters more.

Weekly EEA profile impressions and actions, indexed to Sept 7 to 20Fig 4
0255075100Sep 7Sep 14Sep 21Sep 28Oct 5Oct 12Oct 19Oct 26Nov 2Sept 18Sept 22Directions 93Website clicks 89Maps 75Search 17

Weeks start on Monday.  Google extended the new layout to local searches on Sept 18; Search impressions started falling on Sept 22.  In the week of Sept 28, Search was at 30, Maps at 100, direction requests at 101 and website clicks at 95.  Direction requests and website clicks come from a one-in-six sample of EEA locations.

By early November the gap had widened, and it still sat mostly in website clicks.  Actions fell 7 percent against 2 percent in the same weeks of 2025, about 5 points worse than a normal autumn.  Direction requests, the action most tied to Maps, were within a point of last year, and calls were about 4 points behind.  Website clicks from the profile fell about 9 points more than last year.

Not every lost click is a lost customer.  Someone who books or enquires through a directory in the new block still reaches the business, but Google doesn't count it as a profile action, and it doesn't appear in this data.

Raw declines can look much larger than this, because autumn is slower for most businesses.  Italy's actions fall every year between mid-September and early November, by 9 percent in 2025 and by 14 percent this year, so the change left Italy about 5 points worse off.  That is why every comparison in this brief is set against the same weeks of last year.

Change in customer actions, Sept 7–20 to Oct 26–Nov 8Fig 5
EEA, 2026EEA, same weeks 2025UK, 2026
−10%−5%0%+5%−5%−1%+2%Calls−5%−4%−1%Directions−10%−1%+6%Website clicks−7%−2%+1%All actions

Same locations in both periods.  Direction requests fell by a similar amount in the same weeks of 2025, so most of their late-October decline is seasonal.

If the lost impressions had been costing customers, the businesses that lost the most impressions would also lose the most actions.  They didn't, at least not in any simple way.  Locations that drew most of their impressions from Search lost far more of them than locations that leaned on Maps, yet in every group website clicks ended 5 to 9 points below a normal autumn and direction requests stayed within about 2 points.

By industry the link is loose.  Insurance, financial services and legal lost the largest shares of their impressions and also the most website clicks, more than 20 points against last year.  But automotive, which lost fewer impressions than most industries, still lost 19 points of website clicks, so losing impressions and losing website clicks don't line up neatly.  Restaurants and travel lost the fewest impressions and saw a broader dip, with direction requests about 5 points behind as well.  One possible reason for the professional-services loss is the new directory block: these searches have established comparison and directory sites that may rank first, and clicks inside the block go to the provider.  We can't confirm that yet.

How much worse than last year website clicks and direction requests did, by industry, in percentage pointsFig 6
Website clicksDirection requests
−30−20−100LegalLost 59% of profile impressions−26−1InsuranceLost 70% of profile impressions−220Financial servicesLost 60% of profile impressions−22−3AutomotiveLost 43% of profile impressions−19−2EducationLost 56% of profile impressions−16−2Health and medicineLost 51% of profile impressions−15−1Food and diningLost 39% of profile impressions−9−5TravelLost 37% of profile impressions−9−5ShoppingLost 44% of profile impressions−2+1ContractorsLost 52% of profile impressions−1+4

Each value is this year's change minus the change in the same weeks of 2025, Sept 7–20 to Oct 26–Nov 8.  Under each industry is the share of its Google profile impressions it lost over the same weeks.  A one-in-three sample of EEA locations active in both years, by Yext category.  Legal, education and financial services vary by several points between halves of the sample, shopping and contractors rest heavily on one large client each, and insurance locations are mostly in Germany.

If the change were costing customers in some countries more than others, it would show up in their gaps against last year.  So we compared countries, using only locations active in both years so that clients joining or leaving don't move the figures.  Website clicks fell most in every country except the Netherlands, where they rose.  France stands out as the one large market where direction requests also fell markedly, about 5 points, so something beyond the Search change, such as weaker demand or the later slide in Maps, is at work there.  In the larger markets where we could match industries, mix explains little of the spread, though Italy's weight in automotive accounts for most of its gap.

How much worse than last year actions did, by country, in percentage pointsFig 7
FranceActions −9.4% this year, −2.6% last yearWebsite clicks 9 pts worse, directions 5 pts worse
6.8 worse
SpainActions −11.1% this year, −5.4% last yearWebsite clicks 13 pts worse, directions 3 pts worse
5.7 worse
ItalyActions −13.7% this year, −8.7% last yearWebsite clicks 14 pts worse, directions 2 pts worse
5.0 worse
FinlandActions −6.3% this year, −1.7% last yearWebsite clicks 10 pts worse, directions 3 pts worse
4.6 worse
SwedenActions −1.0% this year, +2.9% last yearWebsite clicks 10 pts worse, directions 1 pt worse
3.9 worse
AustriaActions −1.7% this year, +2.0% last yearWebsite clicks 8 pts worse, directions 3 pts worse
3.7 worse
GermanyActions −2.6% this year, +0.4% last yearWebsite clicks 8 pts worse, directions level
3.0 worse
NetherlandsActions +0.5% this year, −1.4% last yearWebsite clicks 4 pts better, directions 1 pt better
1.9 better

Each value is this year's change in calls, direction requests and website clicks minus the change in the same weeks of 2025.  France's actions fell 9.4 percent this year and 2.6 percent last year, so they did 6.8 points worse.  A one-in-three sample of locations active in both years, countries with at least 1,500 locations in the sample.  France and Germany gave the same result in both halves of the sample; the others moved by 1 to 3 points, so gaps that differ by a point or two are within sampling noise.  The Netherlands' gain rests partly on one large client; without it, the Netherlands was level with last year.

Averages hide a wide range between businesses.  Among accounts with at least 10 EEA locations, comparing each account with its own results a year earlier, about a third did as well as or better than last year on actions, and about a third did 10 or more points worse.  One in four was down 15 percent or more, twice the share of the same weeks in 2025.  A business with a bad number is not an outlier, and the place to start is its own actions against last year.

How to check your own account
  1. Compare actions with last year.  Look at calls, direction requests and website clicks for mid-September and for the latest weeks, against the same weeks of 2025.  A gap of a few points is close to the European average.
  2. Check branded searches in Search Console.  Filter your site's branded queries to EEA countries.  If they are steady, people are still looking for you.
  3. Search like a customer.  Try your main category searches, such as "dentist Lyon", from inside the EEA or through local staff, and note whether a directory block leads the page and which provider is open.
  4. Watch website clicks.  This is where most of the loss shows up.  Some visits that used to come through the profile may now arrive straight from the website carousel, so check visits from Google in your website analytics as well.

The same brands' location pages kept their Search impressions while their profiles lost 87%

If people were searching for these businesses less, or the businesses were being shown less across Google Search in general, we would expect their own web pages in the regular results to follow a similar pattern.  Google Search Console reports how often a website appears in Google results, which gives a second view of the same searches.  Several brands with most of their locations in the EEA have both Google Business Profiles and location pages hosted by Yext, so we compared the two for the same businesses, searchers and weeks.  Their location pages kept their usual Search impressions while their profiles' fell 87 percent, so demand for these brands held.

Before Sept 22 the two moved together.  Afterwards they split.  Between Sept 7 to 20 and Oct 26 to Nov 8, page impressions fell 10 percent, close to the 8 percent decline in the same weeks of 2025, while profile impressions fell 87 percent.

A website result and a profile sit in different parts of the page, and Google has changed where profiles appear on the page, so this comparison does not show how often the profiles themselves were seen.  It does show that demand for these brands held.

Same brands, weekly Google Search impressions: location pages vs profiles, indexed to Sept 7 to 20Fig 8
0255075100Aug 3Aug 17Aug 31Sep 14Sep 28Oct 12Oct 26Nov 2Sept 22Location pages 92Profiles 14

EEA searchers only.  Page impressions come from Google Search Console for the brands' Yext-hosted sites; profile impressions are Google Business Profile Search impressions for the brands' EEA locations.  By late October pages were at 88 and profiles at 13.

If the profiles had really been seen 87 percent less often, their actions would fall far further than clicks on the same brands' location pages, whose visibility held.  So we compared the two week by week.  Profile actions (calls, direction requests and website clicks) and page clicks moved together before and after Sept 22, and both eased in mid-October.  Over the comparison, profile actions fell 10 percent and page clicks 16 percent.  Website clicks from the profiles fell further, mostly from mid-October: 28 percent, about 24 points behind last year against 8 points for page clicks, in line with the website-click losses described above.

Same brands, weekly profile actions and location page clicks, indexed to Sept 7 to 20Fig 9
0255075100Aug 3Aug 17Aug 31Sep 14Sep 28Oct 12Oct 26Nov 2Sept 22Profile actions 92Page clicks 84

Profile actions are calls, direction requests and website clicks from Google Business Profile for the brands' EEA locations; page clicks are Google Search Console clicks from EEA searchers.  The two series have a weekly correlation of 0.90.

People kept searching for brands by name, but profiles counted far fewer of those searches

If the drop came from a change in how Google ranks places, we would expect it to show up in the kinds of searches a business is found for.  In the past, sudden jumps or drops tied to algorithm changes have done exactly that.  When a business suddenly starts appearing for nearby but unrelated searches, such as parking or restaurants, it usually points to a change in how Google Maps ranks places.  So we looked at which search terms lost impressions, to see whether the drop was concentrated in searches for the business's name, category searches or searches unrelated to it.  Every kind of search lost impressions, name searches most, and there was no shift toward unrelated searches of the kind a ranking change usually produces.

Google reports which search terms showed each profile, by month.  Between August, the last full month before the change, and October, impressions from searches for a business's own name fell 57 percent.  In the same months of 2025 they rose 1 percent.  Misspellings and variants of names fell 55 percent.  Category searches such as "restaurants" or "pharmacy" fell 31 percent, about 10 points more than in 2025.

Google's search-term data combines Search and Maps.  Category searches happen mostly in Maps, which changed less, so the 31 percent understates what happened to category searches on the Search results page, which, as the next section shows, is where Google changed its layout.

The same mix explains much of why name searches show the largest drop.  Most name searches happen on the Search results page, where nearly all of the loss occurred, while most category searches happen in Maps, which changed little.  Any kind of search that happens mostly on Search would show a drop of this size.  The harder question is why profile impressions on name searches fell on the Search page at all, since Google's published examples of the new block all involve category searches.

Single-location businesses lost 63 percent of their name-search impressions and chains with 100 or more locations lost 56 percent, so the loss is not limited to chain store maps.  Single locations lose a little more here partly because Google's search-term data leaves out terms with fewer than 16 impressions, which small businesses hit first; in total Search impressions, chains lost slightly more.

Change in search-term impressions, August to October, EEA locationsFig 10
Aug to Oct 2025Aug to Oct 2026
−60%−40%−20%0%+1%−57%Own name−7%−55%Name variants−21%−31%Category searches−14%−30%Other companies' names

Search and Maps combined.  Before the change, own-name searches made up 41 percent of these impressions and category searches 46 percent.  Searches for other companies' names were about 1 percent.

If people were searching for these businesses by name less often, the brands' own websites would lose branded impressions too.  So for the same brands with Yext-hosted pages, we compared counted name searches on their profiles with Search Console impressions on branded queries to their websites, from EEA searchers, for August and October.  Profiles lost 53 percent of their counted name-search impressions, against an 8 percent rise a year earlier.  Branded impressions on the brands' websites fell 5 percent, against a 5 percent rise.

People kept searching for these brands by name.  Profile reporting captured far fewer of those searches, which points to a change in counting as a large part of the name-search loss.  Branded website impressions did slip about 10 points against last year, while those of UK brands held, so some of it may be real, for example on searches that pair a brand with a town and can bring up the new block.

Smaller businesses tell the same story.  In France, plain name searches for single-location businesses lost 68 percent of their counted impressions between August and October, against 10 percent a year earlier, while the address-style searches that Google Maps autocompletes lost 15 percent, against a 10 percent rise.  People did not move their name searches to Maps.

Same brands, name searches August to October: counted profile impressions vs branded website impressionsFig 11
Aug to Oct 2025Aug to Oct 2026
−50%−25%0%+8%−53%Profile name-search impressions+5%−5%Branded website impressions

EEA searchers only.  Profile figures are Google Business Profile search-term impressions containing the brand's name; website figures are Google Search Console impressions on queries Yext classifies as branded, for the brands' Yext-hosted pages.

Of this autumn's changes to the results page, only the new directory block is limited to the EEA

If name and category searches lost counted impressions while people kept searching for these businesses and mostly kept acting on them, something on the results page itself must have changed.  So we looked at which features appear on the page and how they changed, using Semrush snapshots of branded searches and Google's documentation of its new European layout.  AI Overviews and store maps changed in both the EEA and the UK.  The one change limited to the EEA is Google's new directory block.

AI Overviews

AI Overviews appeared on more searches this autumn.  If they were taking impressions that profiles used to get, we would expect profile impressions to fall most where AI Overviews grew most, and to fall when Google switched them on in a new country.  So we tested both, and neither held.  The UK has had them longer, and on more searches, than most of Europe, and its profile impressions did not fall.  Google switched them on in France on July 22, and French profile impressions showed no break.  Across a panel of national store brands in Germany, France, Italy, Spain and the UK, tracked in Semrush, the share of searches showing an AI Overview rose most this autumn in the UK, from 17 percent in August to 49 percent in October.

Share of brand-site searches showing an AI Overview, by countryFig 12 · Semrush
0%25%50%Nov 25DecJan 26FebMarAprMayJunJulAugSepOctFrance launchUK 49%France 26.6%Italy 20.4%Germany 17.4%Spain 11.8%

Average across each country's brands of the share of searches their sites rank for that show an AI Overview, from monthly snapshots on the 15th.  France had none before late July.

Profile panels and store maps

If Google were showing profile panels more often as it showed store maps less, branded searches would shift from one to the other.  So we looked at which features appear together on each brand's top searches.  They did not shift.  The profile panel appeared on 54 to 55 percent of EEA branded searches and 42 to 45 percent of UK ones in both months.  Store maps became less common in both markets, from 35 to 28 percent of branded searches in the EEA and from 49 to 34 percent in the UK, so that decline is not specific to Europe.  AI Overviews took up much of the space, appearing on 31 percent of EEA branded searches in October, up from 5 percent in August, and on 58 percent in the UK, up from 13.

Single-location businesses showed the same pattern.  On name searches for single-location French businesses, Semrush found the profile panel on most of them in November, as in August.

Semrush searches from its own location, which may sit outside the EEA, and may record the new directory block as a store map, so these figures describe broad trends and may differ from the page an EEA searcher sees.

The new directory block

If Google had changed what EEA searchers see, it should show up in Google's own documentation and on the results page itself.  So we read Google's documentation of its new European layout and checked searches in France.  On many local searches, a block of directories now leads the page, and Google's own profiles can sit in a collapsed row beneath it.

On Sept 8 Google introduced two new blocks for searchers in the EEA, first for hotels, flights, trains and products, and on Sept 18 for local businesses.  The first, which Google calls the aggregator unit, shows results from approved directories and booking platforms.  One provider is open at a time, chosen by ranking, and the others are listed beneath it as collapsed rows.  Clicks inside the block go to the provider's own site.  Google's documentation says the block replaces existing units on local searches, without naming them.

Searches checked in France since then show Google's own map and profiles inside the block as one of the providers.  When Google ranks first, profiles appear much as they did before.  When a directory ranks first, a searcher sees its listings, and Google's profiles stay hidden until someone opens that row.  Which directory appears depends on the search; for a dentist or a hairdresser it can be a booking platform.

A second block, the supplier unit, appears only alongside the first.  It is a carousel of business websites built from what Google can crawl, and visits through it reach the business's website directly without passing through the profile.  By Sept 11, Google's documentation listed its older "Places sites" carousel for Türkiye only, so in the EEA the new block appears to have replaced it.

The block does not appear on every local search.  Google has not said how often it does, which provider tends to rank first, or whether profiles in a collapsed row are counted as impressions.

A local category search in France, before and after Google's Sept 18 changeFig 13
Google Business ProfilesDirectory listings and other results
Before Sept 18 boulangerie rennesMore placesWeb results After: a directory ranks first boulangerie rennesAdressesDirectory ABrowse on Directory A ›Directory BGoogleSites Web associésWeb results After: Google ranks first boulangerie rennesAdressesGoogleMore placesDirectory ADirectory BSites Web associésWeb results

Google's map and three profiles lead the page.  Each profile shown logs a Search impression.

The top-ranked directory opens by default and its listings link to its own site.  Google's profiles sit in a collapsed row below it.

When Google ranks first, its profiles open inside the block, much as before.

Simplified sketch, not a screenshot, based on Google Search Central's documented layout for the new units and adapted to a local search: one provider open, the others collapsed beneath it, and a carousel of business websites alongside.  It also draws on searches seen in France in September.  The block does not appear on every local search.  We don't yet know how often each version appears, or whether profiles in a collapsed Google row are counted as impressions.

If the website carousel is showing business sites on category searches, those sites should gain impressions on searches that don't name the brand.  For the same brands, unbranded impressions on their websites from EEA searchers rose 19 percent between August and October, against a 16 percent fall a year earlier, and the rise began the week of Sept 21.  Unbranded impressions for UK brands also improved on last year, by about 13 points, which may reflect citations in AI Overviews, so roughly 20 points of the European rise is specific to the EEA.  Clicks did not rise with them.  It is an early sign that some of the visibility profiles lost has moved to business websites.

Search Console impressions on brands' websites, August to OctoberTable 2
QueriesEEA brands, 2026EEA brands, 2025UK brands, 2026UK brands, 2025
Branded−5%+5%−1%−3%
Unbranded+19%−16%−19%−32%

EEA brands are those whose Yext-hosted pages draw most of their Search Console impressions from EEA searchers, measured on EEA searchers; UK brands likewise for UK searchers.  Branded and unbranded follow Yext's query classification.

How to appear in the new block
  • The directory side.  Each directory in the block builds its listings from its own data, so keep business details complete on the directories that matter in your markets: hours, photos, categories and reviews.
  • The website side.  Google builds the website carousel from what it can crawl, so keep location pages indexable, with clear names, addresses, hours and local business markup.
  • Google's own row.  When Google ranks first, a complete profile matters as much as it did before.

Yext Listings keeps business details consistent across directories, and Yext Pages keeps location pages crawlable and marked up.

Google redesigned EEA results under the Digital Markets Act, with a deadline of about Sept 21

If the drop came from Google's European redesign, its timing should line up with what Google changed and when.  So we lined the two up.  The drop began four days after Google extended its new layout to local searches, and the day after its compliance deadline.

The redesign follows Article 6(5) of the Digital Markets Act, which bars Google from ranking its own services more favourably than similar third-party services.  The July decision found it had favoured its own shopping, hotel, transport and sports results, and Google's redesign extends the same blocks to local businesses, which is why profiles now share the page with directories.

July 23
The European Commission finds Google favoured its own services in Search, fines it €460 million and gives it 60 days to comply.
Sept 8
Google adds two EEA blocks, one for directories and comparison sites and one for the businesses themselves.
By Sept 11
Google's documentation lists the older "Places sites" carousel for Türkiye only.
Sept 18
Google extends both blocks to searches for local businesses.
About Sept 21
The 60-day compliance period ends; missing it risked daily fines.
Sept 22
EEA Search impressions start falling and reach about 30 percent of normal by Sept 28.

Google has not explained what changed in its reporting or how often the new block appears.  The evidence supports two effects working together, and shows what the change cost in customer actions.

Shown less often on category searches
A large share

When a directory ranks first in the new block, Google's profiles sit in a collapsed row and are seen only if the searcher opens it.  This fits Google's documentation, searches checked in France and the drop in website clicks.

Shown but counted differently on name searches
Likely a substantial share

For the same brands, counted name-search impressions on profiles fell 53 percent while branded impressions on their websites fell 5 percent.  Rank trackers still found the profile panel on most name searches we checked, almost no listing fell to zero, and direction requests held.

Effect on customer actions
Small overall, larger in website clicks

The impressions that disappeared were mostly ones that seldom lead to an action.  Actions as a whole were flat in the first two weeks, though website clicks slipped from the start.  Since then the gap has sat mostly in website clicks, about 9 points behind last year and more than 20 in insurance, finance and legal, while direction requests held.

Explanations we tested and ruled out

A data fault
Other publishers reported normally, the drop arrived in two clean steps, and it has lasted two months.
Seasonality
The same weeks of 2025, and the UK this year, show no comparable drop.
Views shifting to Maps
Maps impressions held for three weeks while Search fell 70 percent.
Privacy thresholds
Chains of 100 or more locations lost 84 percent of their Search impressions and single locations 77 percent, the opposite of what thresholds would do, and almost no listing fell to zero.
Something in your listings
Almost every location dropped by a similar share on the same days, so listing content, settings or penalties did not cause it.
AI Overviews
They grew faster in the UK, where profile impressions did not fall.

Four questions we can't answer yet

Each of these would sharpen the picture.  Some need Google to answer, and some need more data or more time.

Do profiles in a collapsed Google row count as impressions?
Needs Google

If they don't, part of the loss on category searches is counting as well, and real exposure fell less than the numbers suggest.  Google has not documented how the new block is reported.

How often does a directory rank first?
Needs a sample of EEA searches

This decides how much real exposure profiles lost on category searches.  A regular sample of local searches run from inside the EEA would estimate it, by country and industry.

Are business websites picking up what profiles lost?
Needs more weeks of data

Unbranded impressions on the same brands' websites rose after Sept 22, but clicks have not followed yet.  A few more months will show whether the website carousel becomes a real source of visits.

Why did website clicks fall most in insurance, finance and legal?
Needs EEA searches by industry

These industries lost more than 20 points of website clicks against last year, more than any others.  If comparison and directory sites tend to rank first in the block for these searches, that would explain it, since clicks inside the block go to the provider.  A sample of EEA searches by industry would show it.

Recommendations

Changes to make in EEA reporting and websites

Measure customer actions

Move EEA scorecards to calls, direction requests and website visits, compared with the same weeks last year, and watch website clicks most closely.  Mark Sept 22 on any chart of Search impressions and set aside year-over-year Search comparisons until September 2027.

Read brand demand in Search Console

In the EEA, name searches in profile reporting no longer track brand demand.  Use branded queries in Search Console instead, where the same brands' branded impressions fell only 5 percent.

Check the directories in the block

Run your main category searches from inside the EEA, note which directories lead the new block, and make sure your listings on them are complete and accurate.  Customers who reach you through them still count, even if Google doesn't report it.  Yext Listings keeps those details consistent.

Keep location pages current

Google's website carousel draws on crawlable pages, so keep location pages accurate and indexable, and tag the profile's website link to see visits Google no longer attributes to the profile.  City and region pages that list three or more locations can carry ItemList and LocalBusiness markup, which makes them eligible for Google's beta location carousel in the EEA.

Editing Google listings will not bring back impressions lost to the new layout.  The places to act are the directories in the new block and your own location pages.

Common questions

Why did Google Business Profile Search impressions drop in Europe?+
From Sept 18, 2026, Google began showing a new block of directories on local searches in the European Economic Area, after an EU ruling under the Digital Markets Act.  Profile Search impressions started falling on Sept 22 and were down 83 percent by early November across 273,000 EEA locations, in every country and industry, while markets outside the EEA held steady.  Part of the drop is real lost exposure on category searches where a directory leads the block, and part appears to be a change in how Google counts impressions on name searches.
Did customers stop finding European businesses?+
Mostly no.  In the first full week after the change, actions as a whole were down 1 percent while Search impressions were down 70 percent, though website clicks had already slipped 5 percent.  Over the full period actions ran about 5 points below a normal autumn, mostly because of website clicks.  Direction requests were within a point of last year.
Did other sites see the same drop?+
No.  Bing's EEA listing impressions rose 13 percent and Tripadvisor's were flat over the same weeks.  EEA directories eased over the autumn, starting the week before, and Apple's engagement figures showed no break on Sept 22.
Are people searching for our brand less?+
Probably not.  For brands with Yext-hosted pages, counted name searches on profiles fell 53 percent between August and October, while branded impressions on the same brands' websites fell 5 percent.  Search Console branded queries are the better measure of brand demand in the EEA now.
Is my business really being shown less?+
On some searches, yes.  When a directory ranks first in the new block, Google's profiles sit in a collapsed row that searchers have to open.  When Google ranks first, profiles appear much as before.  Google has not said how often each happens.
Is it AI Overviews?+
Unlikely.  AI Overviews grew faster in the UK than in the EEA this autumn, and UK profile impressions held steady.  French profile impressions also held steady after Google switched AI Overviews on in France in July.
Should we edit our listings to fix it?+
No.  Almost every location dropped by a similar share on the same days, so listing content, settings or penalties did not cause it.  Keeping location pages and directory listings accurate matters more under the new layout.
Can we compare Search impressions year over year?+
Not for EEA locations until September 2027.  Mark Sept 22, 2026 on charts and compare calls, directions and website visits instead.
Are businesses outside the EU affected?+
Only through searchers inside the EEA.  Swiss businesses near the EEA border lost up to a third of their Search impressions while inland Swiss businesses changed little.  The UK and US held steady.
Which industries lost the most?+
On website clicks, insurance, financial services and legal, which ran more than 20 points behind last year.  Shopping and contractors were close to last year, though each of those results rests heavily on one large client.  Direction requests held in every industry, within about 5 points.
Why did actions slip more in some countries?+
Mostly because of website clicks, which fell in every country except the Netherlands.  France also lost about 5 points of direction requests against last year, more than any other large market, which points to something beyond the Search change.  In the larger markets, differences in industry mix explain little of the spread.  Raw declines also look larger in some markets because autumn is always slower there: Italy's actions fell 9 percent in the same weeks of 2025.

Methodology

Panel
Google performance data for locations managed in Yext: 273,000 in the EEA, with the UK (87,000), Switzerland (55,000) and the US (1.03 million) as comparison markets.  Daily data from Sept 1 to Nov 9, 2026, and monthly data from January 2024.
Comparison
The same locations in both periods, Sept 7 to 20 against Oct 26 to Nov 8, 2026, with the same weeks of 2025 as a seasonal baseline.  Weekly charts index each market to its mid-September level.  Comparisons of actions by country and industry use a one-in-three sample of locations active in both years, so that clients joining or leaving do not move the figures.
New layout
Google Search Central documentation for the aggregator unit, supplier unit, Places sites features and structured data carousels, read on Nov 18, 2026, and searches run in France in September by Yext's French team and in published observations.  Fig 13 is a simplified sketch adapted from Google's documented layout.
Search reliance
Locations grouped by the share of their Google impressions that came from Search in Sept 7 to 20.  A one-in-three sample of EEA locations with at least 100 impressions and 10 actions before the change, compared with the same weeks of 2025.
Other publishers
Listing impressions reported to Yext by Bing, Tripadvisor and EEA directories.  Apple reports place card opens, which measure engagement, so it is described in the text and left out of Table 1.  Google Search Console impressions for Yext-hosted location pages, by searcher country, for brands with most of their locations in the EEA that host their location pages with Yext, compared with the same locations' Google Business Profile Search impressions and actions.
Search terms
Google's monthly search-term data, which covers Search and Maps together and lists only terms with 16 or more impressions in a month.  Terms are grouped by whether they contain a distinctive word from the business's own location names, appear for three or more unrelated businesses, or name another company.  Address-style searches are name searches that include a street or town in the format Google Maps autocompletes.
Search Console queries
Daily Google Search Console impressions and clicks by query and searcher country for Yext-hosted pages, with queries classified by Yext as branded or unbranded.  EEA brands are those whose pages draw most of their impressions from EEA searchers; a comparable set of UK brands serves as a control.  Profile name searches for the same brands are search terms containing a word that appears in at least 30 percent of the brand's EEA location names.
Industries
Each location's Yext category, or its Google category where that is missing, mapped to an industry.  About 113,000 EEA locations could not be matched; their results were in line with the average.  Action comparisons by industry use each location's top-level Yext category and include industries with at least 400 locations in the sample.  The share of profile impressions each industry lost, shown under each industry in Fig 4, comes from the broader matching above, so the two groupings differ slightly.
Semrush
For AI Overviews, monthly counts of the searches each brand's site ranks for that show an AI Overview, a store map or a profile panel, for national store brands in Germany, France, Italy, Spain and the UK, plus the top 25 branded searches for each homepage on Aug 15 and Oct 15, 2026.  For name searches, results pages for single-location French businesses' names on Aug 15 and in November 2026.
Sources
European Commission decision of July 23, 2026 on Google Search under the Digital Markets Act.  Google Search Central pages on the aggregator unit, supplier unit, regional differences in Search experience, Places sites features and structured data carousels, updated between Sept 8 and Nov 9, 2026.
Limitations
Google has not said what changed in its reporting or how often the new block appears, so the split between lost exposure and counting is an inference.  Google's search-term data mixes Search and Maps.  Customer actions count only actions taken on Google, so a call placed through a directory listing is not included, and they do not capture sales.  Semrush checks are small samples from a location that may sit outside the EEA.  A seasonal slowdown affects Maps and actions, and the 2025 baseline removes only part of it.
Yext Research · Google in EuropeBrief 6 · November 2026

Have a question? Let’s study it.

It takes about five minutes. Someone from Yext Research will reply within two weeks.

Pitch a study

Get the latest from Yext Research

See what's changing in search, grounded in real user behavior.

Yext will process your personal data as described in Yext’s Privacy Policy.