We looked at 1,571 owner-operated local businesses across 9 small-business categories and 6 U.S. metros. Seven questions came out of it, each one something an owner could look at in their own setup.
of businesses in the Maps top 20 show up at only one search point across their metro.
reviews: how far the typical business ranked 11–20 sits below its market's rank 4–10 median.
of listings with no reviews never appear in a top 20.
chance a top-3 business out-rated a not-seen rival on stars: a coin flip.
Review count is the attribute most tied to tier, even with distance held equal. We cannot show that gaining reviews moves rank, only that the two go together. Star rating, by contrast, barely separates leaders.
Review count by tier, all categories, one review source for every tier.
CaveatAssociation in one snapshot; order of cause unknown.
Glossaryreview count
datanephew GBP study v1.2 · Maps snapshot 2026-10-18
Median reviews of owner-operated top-3 businesses, by category. Hatched rows are low-confidence and stay explorer-only.
CaveatTwo metros pooled; n per category is small.
Glossaryreview count
datanephew GBP study v1.2 · Maps snapshot 2026-10-18 · also GBP-D1-01 (Low) · also GBP-D1-02 (Low) · also GBP-D1-03 (Low)
Median reviews by category and tier. Color is scaled within each row so the shape of each ladder shows.
CaveatTwo metros pooled; n per category is small.
Glossaryreview count
datanephew GBP study v1.2 · Maps snapshot 2026-10-18 · also GBP-D1-09 (Medium)
Shares of businesses already at the next tier's median review count. Whiskers are 95% intervals.
CaveatTier is best rank across 8 distant points.
Glossaryreview count
datanephew GBP study v1.2 · Maps snapshot 2026-10-18 · also GBP-D2-02 (Medium) · also GBP-D2-03 (Medium) · also GBP-D2-04 (Medium)
Share of listings with no reviews at all, by tier. Ghost listings pile up in the unseen tier.
CaveatRule-based segment; snapshot.
Glossaryreview count
datanephew GBP study v1.2 · Maps snapshot 2026-10-18
Share of owner-operated top-3 appearances with fewer than 10 reviews, by how crowded the search point is.
CaveatThin points concentrate in home health and handyman.
Glossaryreview count, local pack
datanephew GBP study v1.2 · Maps snapshot 2026-10-18
Within-market AUC of star rating, top 3 vs not seen. 0.50 means no separation. The band is the 95% interval.
CaveatRatings cluster at 4.8-5.0 in every tier.
Glossarystar rating
datanephew GBP study v1.2 · Maps snapshot 2026-10-18
Percentile among top-3 and top-10 owner-operators, pooled across metros. Confidence varies by category (see chart 08).
percentile among top 3 accountants
percentile among ranks 4–10 accountants
CaveatNearest search point can be ~9 km from the business.
Glossaryreview count, local pack
datanephew GBP study v1.2 · Maps snapshot 2026-10-18
How many reviews do I have compared with the businesses ranking above me for my category near my address, and how fast am I adding more?
Most businesses appear in one patch of their metro, and rank tends to slip with distance. How fast it slips within a few kilometers is not known yet.
Share of visible owner-operators seen at only one search point across their metro. Whiskers are 95% intervals.
CaveatGrid geometry, not business strength.
Glossaryproximity, local pack
datanephew GBP study v1.2 · Maps snapshot 2026-10-18 · also GBP-C2-01 (High)
Tier mix of listings by distance to the nearest search point, all categories.
CaveatDistance to search points, not to real customers.
Glossaryproximity
datanephew GBP study v1.2 · Maps snapshot 2026-10-18
Share of appearances ranked 1–3, by km from the search point.
CaveatDistance known for 38% of appearances, skewed to better ranks.
Glossaryproximity
datanephew GBP study v1.2 · Maps snapshot 2026-10-18
Share of listings never seen in a top 20, by category and distance. Dashed cells have a small sample (directional).
CaveatDistance to search points, not to real customers.
Glossaryproximity
datanephew GBP study v1.2 · Maps snapshot 2026-10-18
Where do my website visitors come from, and how do I rank for my category from a few different parts of town?
Estate attorneys and handymen who use the exact category show up higher. Pooled across all nine categories the pattern is flat, so it may only apply in some.
Share of listings using the exact primary category, by tier.
CaveatVisible tiers have 13-23 listings each.
Glossaryprimary category
datanephew GBP study v1.2 · Maps snapshot 2026-10-18 · also GBP-E1-04 (Medium) · also GBP-E1-03 (Low)
Almost every top-3 homepage names its metro's city or a street address somewhere in its visible text. The title tab is a weaker signal. These are associations, not effects, and the check does not record where on the page the mention sits.
Where on the homepage the city or an address appears. "Page title" is the title in the browser tab. "Page text" is all the visible text on the homepage, including the header, menu, headings and footer: it counts the metro's main city name or anything shaped like a street address. A suburb's own name does not count. The meta description and schema markup were not checked, and the check does not record which part of the page the mention was in.
CaveatCity match uses the metro name only.
GlossaryNAP consistency
datanephew GBP study v1.2 · Maps snapshot 2026-10-18
Unseen businesses more often list a website that no longer loads, or a phone number that does not match. These look like housekeeping issues. Whether they cost calls is not known.
What happened when we visited each listed website, by tier.
CaveatSmall counts; one DNS check per site.
datanephew GBP study v1.2 · Maps snapshot 2026-10-18
Phone consistency checks, one dot per tier.
CaveatMismatches may be tracking or second-location numbers.
GlossaryNAP consistency
datanephew GBP study v1.2 · Maps snapshot 2026-10-18
Service-area businesses are rarer at the top. It looks like a tradeoff worth understanding, not something to change for its own sake.
Share of visible owner-operators that hide their address, by tier.
CaveatMaps location of service-area businesses is a hidden centroid.
Glossaryservice-area business
datanephew GBP study v1.2 · Maps snapshot 2026-10-18 · also GBP-B3-02 (Medium)
This study is one snapshot. It cannot say whether more reviews come before higher rank, or whether homepage changes bring more calls. Only a business's own numbers over time can. That is the part datanephew can look at.
What changed in my searches, website visits and customer reviews over the last 90 days?
These are accurate and graded, but nothing about them leads an owner to a next step on their own. They stay available for people exploring the data.
Each bar is a stage of the study.
CaveatSnapshot of 18 category-metro pairs, 2026-10-18.
Glossarylocal pack
datanephew GBP study v1.2 · Maps snapshot 2026-10-18
Share of top-3 appearances with fewer than 10 reviews, by category.
CaveatConcentrated in home health and handyman.
Glossaryreview count
datanephew GBP study v1.2 · Maps snapshot 2026-10-18
Tier mix for five rule-based owner profiles.
CaveatSegments are descriptive groupings, not causes.
Glossaryservice-area business
datanephew GBP study v1.2 · Maps snapshot 2026-10-18 · also GBP-E3-04 (Low)
Tier mix by rating band, all categories.
CaveatSmall effect: about 7 percentile points of rank.
Glossarystar rating
datanephew GBP study v1.2 · Maps snapshot 2026-10-18
Odds multiplier of reaching a higher tier per standard deviation. 1 means no association.
CaveatRaw photo counts do rise with tier.
Glossaryphotos, proximity, claimed profile, star rating, review count
datanephew GBP study v1.2 · Maps snapshot 2026-10-18 · also GBP-D5-02 (High) · also GBP-D5-04 (High) · also GBP-D5-05 (High) · also GBP-D5-06 (High) · also GBP-D5-08 (Medium)
Raw gaps by tier, then the adjusted association underneath.
CaveatRaw medians; see GBP-D5-03 for the adjusted view.
Glossaryphotos, claimed profile
datanephew GBP study v1.2 · Maps snapshot 2026-10-18 · also GBP-D5-03 (High) · also GBP-D5-05 (High) · also GBP-E2-02 (High)
Lighthouse mobile score by tier, one run per homepage.
CaveatLab data: one mobile run per homepage; no real-visitor data.
Glossarypage speed
datanephew GBP study v1.2 · Maps snapshot 2026-10-18 · also GBP-G3-02 (High) · also GBP-G3-03 (High)
Share failing the lab rule, and share taking over 5 seconds to show anything, by tier.
CaveatLab thresholds; TBT stands in for INP.
Glossarypage speed
datanephew GBP study v1.2 · Maps snapshot 2026-10-18
Share of homepages with first paint over 5 s, by website platform. Directional.
CaveatDirectional; platform goes with owner type.
datanephew GBP study v1.2 · Maps snapshot 2026-10-18
Raw share of sites with each feature, by tier, sorted by the raw top-3 vs not-seen gap. Raw gaps are not effects.
CaveatHomepage-only checks; 270 rendered sites.
datanephew GBP study v1.2 · Maps snapshot 2026-10-18
These are associations from a single Google Maps snapshot. Nothing here shows that changing a listing changes its rank. Review counts appear only as quantiles of groups of ten or more businesses; no single business is named.
Confidence grades come from the findings registry: High findings pass every robustness check, Medium findings have a clear direction with one known caveat, and Low findings are shown only so nothing is hidden.
datanephew (2026). Google Business Profile local visibility study, data release v1.2 (snapshot 2026-10-18). https://datanephew.com/research/local-pack-study/