Google Maps businesses lose 1.95% of new reviews in 7 days, study finds

Reviewflowz, a review management software company, published a study on October 16, 2026 that followed 4,876,933 Google reviews on 9,048 Google Maps profiles in the United States and Europe every day for two weeks. It found that 1.95% of new reviews were gone within seven days. The research, run with the Paris data and CRM consultancy Cartelis, recorded 4,590 deletions between August 11 and 24, 2026. Its lead finding concerns a rule Google introduced in April: according to the report, reviews that name a person were deleted at 1.01 times the rate of reviews that do not, when compared inside the same business.

In Short

A review software company checked about 9,000 businesses' Google Maps pages every morning for two weeks and counted which reviews vanished. About 2 in every 100 new reviews were gone within a week, and the ones most likely to go were 1-star reviews, reviews from people posting for the first time and reviews on near-perfect profiles, which matters to any shop, clinic or plumber that customers find through Maps. Reviews that mention a staff member by name disappeared no more often than any others, even though Google told businesses in April not to have their staff ask for them, but the study only counted deletions and could not see how any review was requested.

A daily census of 9,048 profiles

Vendor research often draws on the vendor's own customers. This panel was built to avoid that. According to Axel Lavergne, founder of Reviewflowz, the businesses were "picked at random in the US and Europe (not our clients)."

Candidates came from Overture Maps, an open map dataset, according to the report. Locations were grouped into brands by Wikidata identifier, website domain or name, and brands were then drawn at random, each entering with every location it operates in the US and Europe, "so no chain is cherry-picked and no chain is partly sampled." Each location was matched to Google Places by name and address and kept only when the match sat within 100 metres of the map coordinates, with a name similarity above 0.8, on an operational business. A brand qualified only if its first resolved location carried between 100 and 10,000 reviews.

The panel splits almost evenly by region: 4,508 profiles in the United States and 4,540 across 40 European countries, with the United Kingdom excluded. France contributed 961 locations, Germany 904, Spain 551 and Italy 522. Seven industries are represented - automotive, food and beverage, healthcare, home services, hospitality, travel, and wellness and fitness - across three group sizes: single locations, groups of 4 to 10 and groups of 20 to 50, with about 238 profiles per cell. Groups of 2 to 3 and 11 to 19 locations were left out to keep the size buckets clean, and chains above 50 locations were excluded. Travel has single locations only, because groups of travel operators were too rare to fill the larger cells. The median profile carried 231 reviews and the mean 532.

Every review on every profile was collected each morning from August 11 to 24, 2026, for 14 complete crawls. A deletion is a review present in one complete listing and absent from the next complete one. Crawls that did not reach the end of a listing were discarded, "so a missing page is never mistaken for a deletion," the report states.

Two populations were followed. New reviews are the 25,895 reviews published from August 7 to 17, each tracked for seven days. Existing reviews are every review online on August 11, tracked through August 24.

Six brands were set apart. Four US pest-control chains with 102 locations and two Spanish gyms were hit by mass removals during the study, accounting for 1,568 deletions, about a third of the total. Left in, "they would drive every average," according to the report. With those brands and three profiles that dropped out removed, the base for most figures is 8,890 businesses, 4.59 million reviews and 3,022 deletions.

The statistics rely on Poisson regressions with one term per business, which compare reviews sitting on the same profile, and logistic regressions for first-week outcomes, with errors clustered by business and 95% intervals reported throughout. Cross-business comparisons add controls one group at a time, and brand-level results use a brand bootstrap because purges hit whole brands. The analysis ran in Python, BigQuery and DuckDB, and the report says the scripts and aggregated tables are available on request.

The staff-name rule and the 1.01x result

The study was framed around a policy change. Since April 2026, Google's review policy has said businesses must not direct staff to request reviews that include specific content, "including content that identifies a staff member," according to the report. PPC Land reported the change on April 17, 2026, when two clauses were added to the Rating Manipulation section of Google's Maps user-generated content policy, barring merchants from asking staff to solicit a set number of reviews or reviews with specific content. Customers naming staff on their own remains allowed.

How a review was requested cannot be seen from outside. Whether reviews that name people survive can. To find them, Reviewflowz and Cartelis ran a named entity recognition model called GLiNER over 2,675,861 positive reviews with text. The detection was validated on 600 reviews and then tested blind on 1,000, where it produced 0.66% false detections and found 93% of the names present, according to the report.

Names turn out to be common. 28% of positive reviews name a person: 38% in the US and 13% in Europe, from about 7% in hospitality to 66% in US home services. Three quarters of those names recur in three or more reviews of the same business, which the report takes as evidence that they belong to staff.

In raw numbers, named reviews look riskier, deleted 1.32 times as often as unnamed ones. The report attributes that to where they sit: in businesses and periods that lose more reviews anyway. The comparison it treats as fair holds the business, the review's age and its length constant. On that basis the ratio is 1.01x, with a 95% interval of 0.91x to 1.11x, which rules out any increase above 11%. The test was run fifteen ways, twelve of which are published. None produced a significant result.

Deletion rate of reviews naming a person against reviews that do not; 1.0x means no difference. Source: Reviewflowz and Cartelis.

The pest-control figure is the most striking. In the largest purge the study observed, four chains lost 1,199 positive reviews in two weeks, and named reviews were removed slightly less often than unnamed ones.

"Google does not enforce it," the report concludes about the April rule. Lavergne puts it the same way: "Google's April rule against reviews that name staff isn't enforced."

What a deletion count can and cannot show

That conclusion rests on a single enforcement channel. The study measures reviews that were published and later removed. It cannot observe reviews held back before publication, and it does not track restrictions placed on business profiles. Google's help documentation on review requests, formalized at the end of 2025, states that fake engagement can lead to restrictions on a Google Business Profile. Google's 2025 figures, published in April 2026, included posting restrictions on more than 782,000 policy-violating accounts.

The April rule also targets a business practice - directing staff to solicit named reviews - rather than the content of a review. A customer who mentions a technician unprompted breaks no rule, so a blanket deletion of named reviews would sweep up compliant reviews along with the rest. The report acknowledges the boundary in its limits section: "The names results describe what Google does with named reviews, not whether a given business asked for them." The report contains no comment from Google.

Name-heavy businesses lose more, then collect more

At the level of whole businesses the picture is less tidy. Businesses where half or more of positive reviews name someone - name-heavy businesses, in the report's terms - lost 0.13% of their reviews over two weeks, against 0.03% for businesses where under 10% do. That is 13 reviews in every 10,000 against 3, a 3.9x gap (2.7x to 5.7x).

Half of the gap is sector, according to the report: home services, healthcare and automotive, where naming the technician is normal, are also where the most reviews disappear. At equal sector, region and group size, the gap narrows to 2.4x (1.6x to 3.8x). The rest is pace and freshness. Name-heavy businesses collect about 10 reviews a month against 5, their reviews are twice as young, and a review under two weeks old is about 100 times as likely to be deleted as one over three years old. At equal volume, pace and freshness, the ratio drops to 1.26x, with an interval of 0.79x to 2.01x, and the report describes the difference as no longer measurable. That interval is wide. It does not show a penalty, but it does not exclude one of up to about double at the business level either.

Part of the residual sits in the US, where 6.4% of name-heavy businesses lost at least one recent positive review against 1.3% of businesses under 10% - roughly 4.9 times as many, by PPC Land's calculation. About half of that gap is their higher review volume, according to the report. None lost all its recent reviews, and only 8 lost a quarter or more. In Europe the difference, 3.3% against 2.3%, was not measurable. When a name-heavy business was hit, named and unnamed reviews went alike (0.98x). "Nothing in the reviews themselves explains which businesses were hit," the report states.

The report then sets out what it presents as the value of names. At equal sector, region and size, name-heavy businesses gain 2.7 times as many reviews a year (2.2x to 3.2x), lose 2.4 times as many and finish the year with twice the net reviews.

Averages per business. Gained: reviews from the twelve months before the study. Lost: deletions during the two weeks, scaled to a year. Source: Reviewflowz and Cartelis.

The report's own text carries three caveats on these figures. The yearly losses multiply two weeks in August by 26. The typical business in the name-heavy group collects 30% to 80% more reviews, not 2.7 times as many, so the average is pulled up by heavy collectors. And the 25% to 50% band, not the most name-heavy group, loses the most per 100 reviews gained. The report also found that a business whose own reviews start naming people collects more in those months: +36% year on year for the median business whose name share rose by 25 points. Healthcare runs the other way, with name-heavy businesses collecting fewer reviews. All of these are associations. A business that trains staff to ask for reviews is likely to collect both more reviews and more named ones, and nothing in the data separates the two.

A seven-day probation

Timing produces the clearest pattern. The report describes a new review as "on probation for seven days." Measured as the share of reviews still online that were deleted on each day of their life, the risk is low on day 1, rises on days 2 and 3, dips, then peaks on days 6 and 7. After day 8 it falls to about 0.1% a day, a tenth of the day-7 level.

Negative and positive reviews are removed at different moments. On days 2 and 3, 1.39% and 1.33% of 1- and 2-star reviews still online were deleted. Positive reviews went later: 0.65% on day 6 and 1.03% on day 7. Negatives get a second pass at that point, at 0.88% and 1.58%.

Share of reviews still online deleted that day. Fewer reviews are followed on later days because reviews posted late in the study have fewer days left to observe. Source: Reviewflowz and Cartelis.

Is this a weekly clean-up? According to the report, no. Reviews posted on a Monday peak on day 6 and those posted on a Friday on day 7 or 8, so the peak follows the age of the review rather than the day of the week - "an age trigger, not a weekly batch." Without the six brands, 47% of the positive reviews deleted in their first 13 days went on day 6 or 7. With them, the share is 56%.

By day 7, 1.69% of new positive reviews were gone, rising to 2.50% by the end of the study. For negative reviews the figures were 4.73% and 6.12%. "A campaign looks better on day 1 than it is," the report states.

The extremes go first

By star rating, deletions trace a U-shape. In their first week, 5.7% of new 1-star reviews disappeared, against 2.2% of 2-star, 0.85% of 3-star, 0.97% of 4-star and 1.8% of 5-star reviews.

Inside the same business and at the same age, a 1-star review was deleted 3.9 times as often as a 4-star one (3.1x to 5.0x), a 2-star 2.9 times and a 3-star 1.9 times. The 5-star ratio of 1.15x (0.95x to 1.38x) falls within the margin of error, and the report attributes the extra 5-star deletions in the raw figures to which businesses get hit rather than to the rating. Reviews of 1 and 2 stars make up 9% of all reviews and 20% of deletions. In the US, 6.3% of new negative reviews were gone within a week; in US automotive, 10.2%.

The text itself barely registered. Review length made no difference inside a business (1.04x). Phrases such as "highly recommend" look twice as risky in raw numbers and make no difference inside a business (0.97x): according to the report, "they mark solicited reviews, they do not trigger deletions." Reviews with a photo were deleted less often in the first week, 1.1% against 1.8%, but with controls the gap shrank to between 0.76x and 0.83x and stayed within the margin of error.

What about ratings? Because more 1-star than 5-star reviews disappear, the effects largely cancel. Among the 116 businesses that lost five or more reviews, the average rating moved by 0.01 star.

The account matters more than the text

"Who wrote the review matters more than what it says," according to the report. Accounts posting their first review wrote 14% of reviews and accounted for 22% of deletions. Over the two weeks, 0.10% of existing first reviews were deleted, against 0.02% for reviews by Local Guides at level 6 or above, a 5.6x difference in raw terms. Inside the same business, a first review was 1.44 times as likely to go as one from a reviewer with 21 or more reviews (1.22x to 1.70x), and a review from a Local Guide at level 6 or above 0.59 times as likely as one from levels 1 to 3 (0.50x to 0.70x).

Source: Reviewflowz and Cartelis.

Some deletions are not judgements on a review at all. Among 334 deleted reviews whose author had other reviews in the panel, 64 (19%) disappeared together with every other review by that author, 52 of them on the same day across different businesses. The report reads these as Google removing an account. Only 0.1% of authors with two or more reviews in the panel lost any review. Account removals also account for some of the old reviews that vanished: 39 of the 958 deleted reviews older than a year went together with every other review by their author.

Replies and spike days

Owner behaviour showed up in two places. Among new 5-star reviews, 1.5% of those that received an owner reply within two days were deleted between day 3 and day 7, against 2.4% of those without one. At equal sector and size the odds were roughly halved (0.51x, 0.31x to 0.84x). Within those two days timing made no difference: same day 0.46x, next day 0.52x, day 2 0.51x. Across the whole stock, inside the same business, a replied review was deleted 0.72 times as often (0.60x to 0.87x).

Habit did not count. Businesses replying to three quarters or more of their reviews fared no differently from those replying to none (1.0x); the reply on the specific review is what correlates with survival. On 1-star reviews, replies made no measurable difference (4.9% against 6.0%). The report labels the finding "association, not proven cause." Later replies cannot be compared fairly, because a review has to survive to receive one, and neither the content of replies nor any flags were observed.

Volume on a single day mattered more than volume over a month. The report defines a spike day as one on which a business receives at least 3 positive reviews and at least 4 times its daily average of the previous year. Reviews posted on such days were deleted within a week 2.3% of the time, against 1.5% on other days, or 1.8x at equal sector and size (1.1x to 2.8x). The effect was strongest in Europe, at 2.2% against 0.9%, or 2.75x. In the US, the gap of 2.4% against 2.1% fell within the margin of error.

Risk rose with the count on the day: 1.15% for a lone positive review, 1.65% for two, 1.89% for three or four and 3.08% for five to nine. Businesses receiving ten or more a day, typically large chains with a high natural flow, came in at 1.79%. Monthly surges behaved differently. Businesses whose monthly volume tripled lost 1.1% of their recent reviews, against 1.9% in normal months. "Google reacts to the day, not the month," according to the report.

That finding intersects with Google's own guidance. When Google documented its review link and QR code tool on December 31, 2025, it listed receipts, thank-you emails, the end of chat interactions and printed in-store QR codes as places to share review requests. A request sent to a whole customer list on one day can produce exactly the pattern the report defines as a spike.

Near-perfect profiles

Profiles dominated by 5-star reviews lost more. Businesses where 85% to 95% of reviews are 5-star lost 0.13% of their reviews over two weeks, and those at 95% or more lost 0.14%, against 0.03% for businesses at 50% to 70%. At equal sector, size and volume, the ratios against the 50% to 70% group were 1.4x for the 70% to 85% band and 2.2x for each of the two top bands. In profiles rated 4.8 and above, even a 4-star review was deleted 9 times as often as in profiles rated under 4.5. With the most-hit 1% of businesses set aside, the gap held at 1.5x to 1.7x.

"A 4.9 built on 97% five-stars is a risk profile," the report states.

Where the deletions land

Sector and geography split sharply. In the first week, 3.9% of new positive reviews in home services were deleted, against 2.6% in healthcare, 2.1% in wellness and fitness, 1.8% in travel, 1.3% in automotive, 0.85% in food and beverage and 0.75% in hospitality. At equal region and size, home services lost 3.3 times and healthcare 2.6 times as much as automotive.

Share of new positive reviews deleted within 7 days, reviews posted August 7 to 17. Source: Reviewflowz and Cartelis.

The US lost almost twice as much as Europe: 2.1% of new positive reviews within a week against 1.2%, and 6.3% of new negative reviews against 3.6%. On the existing stock, 0.09% of US reviews were removed over two weeks against 0.04% in Europe. US groups of 4 to 10 locations were the most exposed segment at 0.11%, and US home services groups of that size reached 0.34%. Spain stood out in Europe, with 2.9% of new reviews deleted within a week, ahead of the US at 2.4%, Germany at 1.8% and France at 1.4%. Italy (0.9%), Romania and Portugal (0.4% each) were lowest. Countries with fewer reviews are less precise, the report cautions, because "one purged business can move them."

Deletions were concentrated. 13% of businesses lost at least one review in two weeks - 12% in Europe and 14% in the US - and 810 lost exactly one. The top 1% of businesses accounted for half of all deletions and the top 5% for three quarters. Exposure grew with volume: 4% of businesses collecting under two reviews a month were hit, against 67% of those collecting 50 or more. A business hit in the first week had a 33% chance of being hit again in the second, against 6.5% for the rest.

Purges, review bombs and old negatives

The six excluded brands show what mass removals look like. Two Spanish gyms received 297 1-star reviews on August 1 and 2; 295 were gone by August 16. A German bus tour operator with 156 reviews lost 46 1-star reviews, posted between 2018 and 2026, in four days. Four US pest-control chains lost 1,199 positive reviews across 102 locations - 26% of all deletions in the panel - in purges on August 12, 16 and 17 and a steady removal of fresh reviews, with 310 going one week after they were posted. A US balloon operator lost 38 5-star reviews in three days on a profile of 1,295. Two of the 9,048 profiles were delisted entirely.

Daily counts on the Reviewflowz research page trace the same events: 481 deletions on Wednesday, August 12, the first pest-chain purge; 657 on Sunday, August 16; 750 on Monday, August 17, the largest single day; 266 on Wednesday, August 19, when two profiles were delisted; and 476 on Sunday, August 23.

The gym episode resembles a process Google described in April. When a spike in spam reviews gets through, Google said it removes the fake content, pauses new reviews on the profile, alerts the owner and shows consumers a banner. The removal of years-old negatives has a precedent in Local Services Ads. When Google moved review management for those ads to Business Profiles on July 11, 2025, it said all existing reviews would be verified against Google Maps review policies, a check that reached back to historical reviews.

Purges distort averages, which is why the six brands were held out. With the pest-control chains included, home services appeared to lose 9.8% of new reviews against 1.0% for restaurants; without them, 3.9% against 0.85%. Including all six brands lifts the share of new reviews deleted within a week from 1.95% to 2.76%. "Figures that mix purges with ordinary deletions overstate everything," according to the report.

The overall rate, set against Google's numbers

How large is 1.95%? Over the two weeks, removals of reviews of every age equalled 9.9% of the new reviews published, or 13.9% with the six brands. For positive reviews that ratio was 11.6% in the US and 5.3% in Europe; for negative reviews, 22.9% and 19.0%. The report stresses that this compares deletions of all ages with new publications and is not the share of new reviews deleted.

Existing reviews erode slowly. During the study, 0.055% of the reviews online on August 11 were removed. At that pace a profile would lose about 1.4% of its reviews a year, and about 0.75% of those older than three months. Half of all deletions (48.6%) hit reviews under 30 days old, and a third hit reviews over a year old. The median deleted review was about a month old.

Deletions over two weeks, by the age of the review. Source: Reviewflowz and Cartelis.

These rates sit far below the totals Google publishes, and they measure something different. In its 2025 Trust and Safety Report, Google said it blocked or removed more than 292 million policy-violating reviews out of more than 1 billion published, roughly 22% of review activity. That figure combines reviews blocked before publication with those removed afterwards. A daily crawl of live profiles can only see the second group.

Limits the report sets out

The authors list their own constraints. The window was two weeks in August, so seasonality and longer cycles are not covered; the report says "a second wave will test them." The results are associations: flags, reports, legal requests and authors deleting their own reviews are all invisible to the crawl and counted as deletions alike. "We see what was deleted, not why," the report states.

The crawl looked once a day. A deletion is dated to the morning after it happened, and a review posted and removed on the same day is never seen. Reviews removed before the first crawl on August 11 were also missed, so first-week rates for reviews posted from August 7 to 10 are a floor. The panel covers seven sectors, the US and Europe without the UK, businesses whose first location had 100 to 10,000 reviews and three group-size bands, and results may differ for very small profiles, the UK and other sectors. Finally, the authors note that with dozens of comparisons, a few will look significant by chance, and they label single subgroup findings as exploratory.

Who ran it, and what it sells

Reviewflowz sells review management software. According to its description in the report, the product sends review requests triggered by sales recorded in Stripe, Square, Shopify, HubSpot, Salesforce or Clio, routes each customer "to the platform that needs new reviews most," monitors more than 200 review platforms for new reviews, flipped ratings and deleted reviews, answers reviews through agents trained on a business's own reply history, and attributes reviews to individual team members, "nicknames and typos included, with team leaderboards." The report states its motive directly: "Reviewflowz catches deleted reviews as they happen. This study measures how often Google deletes them, and why." Cartelis is described as a Paris consultancy working on CRM and relationship marketing, analytics, data engineering, customer data platforms and AI.

That commercial context shows in the recommendations. The report closes with five recommendations for keeping reviews online: judging campaigns on day 8, replying to every positive review within two days, never letting a campaign land on one day, asking customers who already review on Google Maps, and asking customers to name who served them. The fifth runs against the conduct Google's April policy describes. "Google's rule says not to, but it does not enforce it," the report states, and staff attribution is one of the features Reviewflowz sells. Elsewhere, drawing on the first-review results, the report argues that asking customers to create a Google account just to leave a review serves little purpose and that sending them to another platform is often the better option - a routing decision Reviewflowz's software automates.

Why this matters for marketers

Reviews are a ranking input, not decoration. PPC Land's analysis of how Google Business Profile determines local search rankings placed review volume and quality under prominence, one of the factors Google uses to order local results. Ratings also travel beyond Maps: since Gemini began serving local results from Google Maps in a visual format in December 2025, those cards display Maps ratings alongside photos. A review that disappears on day 7 drops out of every surface that draws on the profile.

For advertisers, the sector split matters as much as the averages. Home services, the category with the highest first-week deletion rate in the US at 4.1%, covers the plumbers, electricians and locksmiths that buy Local Services Ads, and since July 2025 the reviews shown on those ads have been managed from Business Profiles. Agencies that report review growth to clients are measuring a number that, on this evidence, keeps moving for at least eight days after a review appears, and that concentrates its losses on a small share of accounts: the top 1% of businesses carried half of all deletions.

The policy question remains open. The data shows that, over two weeks in August, Google did not remove reviews naming staff more often than other reviews on the same profiles. It does not show whether Google acts on the April rule through profile restrictions, pre-publication filtering or account-level action, none of which a crawl of published reviews can observe. Nor does it show whether businesses that directed staff to collect named reviews faced consequences, since the study cannot tell which businesses did so.

Timeline

• November 17, 2024: PPC Land details how Google Business Profile determines local rankings, with review volume and quality under prominence.
• March 2025: Google's review link and QR code tool appears in Business Profile without official guidance.
• July 11, 2025: Google moves Local Services Ads review management to Business Profiles and says existing reviews will be checked against Maps policies.
• December 13, 2025: Gemini starts showing local results from Google Maps as visual cards with ratings.
• December 31, 2025: Google publishes its help page on review requests, listing fake engagement as grounds for profile restrictions.
• April 16, 2026: Google reports more than 292 million reviews blocked or removed in 2025 and describes its response to spam spikes.
• April 17, 2026: Google's Maps policy bars businesses from directing staff to solicit review quotas or reviews naming staff.
• August 1-2, 2026: Two Spanish gyms receive 297 1-star reviews.
• August 7, 2026: Start of the publication window for new reviews in the study.
• August 11, 2026: First daily crawl of 9,048 Google Maps profiles.
• August 12, 2026: First pest-control chain purge, 481 deletions that day.
• August 16, 2026: 295 of the 297 gym reviews are gone; 657 deletions that day.
• August 17, 2026: Largest single day, 750 deletions; end of the new-review window.
• August 19, 2026: Two profiles delisted.
• August 24, 2026: Final crawl; 4,590 deletions recorded in total.
• October 16, 2026: Reviewflowz and Cartelis publish the study.

Related PPC Land coverage

• Google tightens Maps review policy: staff names and quotas now banned - the April 2026 clauses on staff quotas and staff-named reviews that the study set out to test.
• Google Maps uses Gemini to stop fake edits and blocks review extortion scams - Google's 2025 enforcement figures and its process for pausing reviews after spam spikes.
• Google formalizes review requests nine months after feature launch - the help page for review links and QR codes, including the fake engagement rules.
• Google shifts Local Service Ads reviews to Business Profiles management - the July 2025 move that put historical Local Services Ads reviews through a Maps policy check.
• How Google Business Profile determines local search rankings - the ranking factors, including reviews as part of prominence.
• Gemini now serves local results from Google Maps in visual format - how Maps ratings now appear inside Gemini answers.

Summary

Who: Reviewflowz, a review management software company, with Cartelis, a Paris data and CRM consultancy. Founder Axel Lavergne says the sampled businesses were not Reviewflowz clients. Affected parties include local businesses, agencies and advertisers relying on Google Maps reviews.

What: A study of 4,876,933 reviews on 9,048 Google Maps profiles crawled daily for two weeks. It recorded 4,590 deletions and found that 1.95% of new reviews were gone within seven days, that 1-star reviews, first reviews, reviews posted on spike days and reviews on near-perfect profiles were removed most often, and that reviews naming a person were deleted at 1.01 times the rate of other reviews on the same profile.

When: Data collected from August 11 to 24, 2026, covering new reviews posted from August 7 to 17. The study was published on October 16, 2026.

Where: 4,508 profiles in the United States and 4,540 across 40 European countries, excluding the United Kingdom, in seven industries.

Why: Google added rules in April 2026 barring businesses from directing staff to solicit reviews that name staff members. The study tested whether such reviews are deleted more often. It found no difference in deletions, but it cannot observe profile restrictions, pre-publication filtering or how any review was requested, so it does not settle whether Google enforces the rule through other means.