MapsSEO

UK Google Maps study: reviews and changes at the top

We collected 5,952 listing appearances from 300 Google Maps searches on 6 July 2026. We used 15 search terms at fixed points in 20 UK cities and repeated the searches on 7 July. A business could appear in more than one search.

MapsSEO research · Method and figures updated

Of 300 repeat searches, 203 kept the same three names in order, 34 only reordered them and 63 had at least one name change.

In short

In the first capture, listings with a recorded review count had a median of 88 reviews among the first three stored results and 57 in the later results. The next day, 63 of 300 first-three lists had a different combination of displayed names. These figures describe what appeared at the sampled points on those dates.

Our UK search-term comparison examines a different question. It compares the first three stored names for a trade and a specific job at the same city point on the same day.

If you are choosing an area for a new branch, start with the premises your business can serve. Our comparison of registered business and hotel sites in six council areas includes the official counts and a downloadable table.

What the first capture showed

We compare the first three stored Maps results in each search (900 appearances) with the remaining 5,052. Reviews and ratings use only rows with a value recorded for that measure.

88 vs 57

Median recorded reviews

Calculated from 848 first-three appearances and 4,635 later appearances with a review count.

39.4%

Fewer than 50 reviews

334 of 848 first-three appearances with a recorded review count fell below 50.

7.4%

A rating below 4.0 stars

63 of 848 rated appearances among the first three had fewer than four stars.

91.4% vs 86.7%

Marked as claimed

First three compared with later results. A website link was recorded on 86.8% of first-three appearances and 85.5% of later appearances.

How review counts varied by search term

Each row compares the median review count in the first three stored results (top bar) and in the remaining results (bottom bar). The ratio divides the first median by the second. The rating column shows the mean of recorded ratings among the first three.

Query Review medians Ratio Mean rating
Indian restaurant
First three: 765
Later results: 527.5
1.5× 4.7
restaurant
First three: 636.5
Later results: 789
0.8× 4.7
dentist
First three: 253
Later results: 158
1.6× 4.7
dental implants
First three: 251
Later results: 163.5
1.5× 4.7
hair salon
First three: 162
Later results: 96
1.7× 4.8
MOT test
First three: 100.5
Later results: 116
0.9× 4.6
conveyancing solicitor
First three: 87.5
Later results: 75
1.2× 4.5
solicitor
First three: 83
Later results: 35
2.4× 4.5
plumber
First three: 75.5
Later results: 30.5
2.5× 4.8
locksmith
First three: 63
Later results: 40
1.6× 4.7
emergency plumber
First three: 55
Later results: 34
1.6× 4.8
electrician
First three: 49.5
Later results: 17
2.9× 4.9
roofer
First three: 27.5
Later results: 15
1.8× 4.8
accountant
First three: 20.5
Later results: 7
2.9× 4.8
EV charger installation
First three: 10
Later results: 2
5× 3.7

For “restaurant”, the median was 636.5 reviews among the first three and 789 among later results. Read the comparison for the search term you work with. The overall median combines all 15 terms and can hide differences between them.

Paid search cost per click, by query

These Google Ads figures use search phrases that include a city name. The archived cost per click ranges from US$0.23 for “indian restaurant nottingham” to US$48.10 for “emergency plumber london”. They describe paid clicks. We collected no revenue, lead or organic click values in this study.

The figures below show median cost per click for each term. We had a CPC value for 295 of 300 city keywords. The 5 without a value were excluded.

locksmith US$24.49
emergency plumber US$15.28
conveyancing solicitor US$14.96
accountant US$12.31
dental implants US$12.06
solicitor US$8.71
plumber US$6.12
dentist US$5.98
electrician US$4.71
ev charger installation US$3.95
roofer US$3.58
mot test US$3.51
hair salon US$0.89
restaurant US$0.65
indian restaurant US$0.44

How the first three results changed overnight

We repeated the same 300 search-term and city pairs an average of 24.2 hours later. We compared the exact displayed names, counting each occurrence if a name appeared twice.

63 of 300

A change in names

The second list had a name occurrence the first list could not match. Order does not affect this comparison.

97 of 300

A different ordered list

At least one corresponding position showed a different name. In 34 cases the same names simply changed places.

67 of 900

Unmatched name occurrences

Second-day entries that could not be paired with the same displayed name in the corresponding first-day list.

We did not store stable profile IDs. Branches can share a title and a profile can change its name, so these figures track displayed names and their order.

How we collected and counted the results

Collection

We requested 20 results at zoom level 14 using one fixed city-centre coordinate per city. We searched in English for terms such as “plumber”, “emergency plumber” and “dental implants”. Each pass used the same terms and coordinates. The first pass retained 5,952 Maps listing appearances, and the repeat retained 5,951.

Missing fields

The first pass had 52 missing review counts among the first three results and 417 in later results. We exclude these rows from the review calculations. A recorded zero would be a value in that calculation. Claim status and website-link presence describe the fields recorded at capture.

The 20 cities

London, Birmingham, Manchester, Leeds, Glasgow, Liverpool, Bristol, Sheffield, Edinburgh, Cardiff, Newcastle, Nottingham, Leicester, Southampton, Portsmouth, Brighton, Coventry, Bradford, Stoke-on-Trent and Hull.

Absolute rank and stored order

The archive kept listing rows in the order returned. Each row also has an absolute rank from the original response, which can skip positions after other result types are filtered out. In the first pass, the first three stored rows spanned absolute ranks 1–5 and the later rows spanned 4–22.

What the November update changes

The July version grouped the 844 listing appearances at absolute ranks 1–3 and the 5,108 at absolute ranks 4–22. It also counted 46 missing review counts as zero in the under-50 figure and labelled new names as position changes. This version compares the first three stored rows with the later rows, excludes missing review counts and reports both changes in names and changes in order.

The audit file provides aggregate counts, denominators and definitions for both methods.

Download the July 2026 method audit (JSON)

Zoom into your city

Read our city breakdowns of the July 2026 captures. Each looks at a different local question, from review counts to how much the first results changed on a repeat search.

Scope of the findings

  • One point per city. A result from a city centre answers a location-specific question. A business checking its service area needs observations from the places customers actually search.
  • Two July captures. The differences describe one overnight interval for 15 English search terms. A longer series would be needed to measure weekly or seasonal changes.
  • Google Maps results. The sample contains Maps listing rows. Google's main search results and its local three-result box were outside the capture.
  • Observed associations. The comparison describes the listings we saw. Establishing the effect of a profile change on rankings would require a different study.

See where you rank in your own patch

Tell us your service and postcode. We’ll check how your business appears on Google Maps across your service area.

✓ Free · ✓ No account needed · ✓ Checked by a human

Results by email, plus occasional UK local SEO tips — unsubscribe anytime. Privacy Policy