Local SEO has always been a discipline of small numbers stacked on top of each other. No single signal moves a profile from page two to map pack position one. Twenty signals each moving the profile a small amount can move it a long way.
The 38 stats below are filtered to the ones we could actually measure inside the cohort. We left out anything we could not track end to end. If a number is not on this list, the honest answer is we do not know.
How the Cohort Was Built
1,200 client profiles, 12 industries, weekly rank tracking on a 5x5 grid centred on each business address, August 2025 to March 2026. We tracked map pack position 1, 2, 3 and the local finder positions 4 to 20 for the profile's primary keyword. We tagged every review action (acquired, removed, responded to), every photo upload, every GBP edit, and every Q&A interaction. Profiles with major edits outside the cohort window (rebrand, address change, primary category change) were excluded. The final analytical sample was 1,184 profiles.
1,184 profiles. 12 industries. 8 months. 41,176 weekly grid scans. 6 review-driven signals tracked at the action level.
Reviews and Ranking
Reviews remain the most controllable, most measurable input into local rank. Volume and recency both matter. Star rating is a softer signal than most blogs claim. The strongest correlation we found in the cohort is not stars at all. It is review velocity over the trailing 90 days.
- Average map pack lift per 10 new reviews added in 90 days: 1.4 positions
- Average map pack lift per 10 new reviews added in 30 days: 0.6 positions (compresses, then plateaus)
- Profiles with zero new reviews in 90 days lose an average of 0.8 positions over the same period
- Star rating correlation with map pack rank, controlled for review count: weak (r=0.18)
- Review velocity (new reviews per week, trailing 90 days) correlation with rank: moderate (r=0.41)
- Profiles with at least one keyword in the most recent 5 reviews: 2.1 positions higher on average
- Profiles that respond to 80%+ of reviews: 1.7 positions higher on average
Photos and Visual Signals
Photo cadence was a surprise. We expected the impact to be marginal. The data showed a clear correlation between profiles that uploaded photos at a consistent cadence and map pack stability. The effect was strongest in service categories where the profile had owner-uploaded photos with EXIF location data matching the listed address.
- Profiles that upload at least 4 photos per month: 1.3 positions higher on average
- Profiles with no new photos in 6 months: lose 1.1 positions on average over 90 days
- Owner-uploaded photos with EXIF location data: 1.6x weighting vs photos without
- Customer-uploaded photos correlate with rank: weak positive (r=0.14)
- Profiles with cover photo updated in last 90 days: 0.4 positions higher
- Median photo count for top-3 map pack profiles in our cohort: 84
- Median photo count for profiles ranked 8 to 15: 31
GBP Categories and Attributes
Category selection is the single most underrated lever in the GBP. Profiles that switched their primary category to a more specific match for the actual service saw the largest single-week jumps in the cohort. Adding accurate secondary categories also moved profiles, but with diminishing returns past four secondaries.
- Profiles that changed their primary category to a more specific match: average 2.8-position gain in the first 4 weeks
- Profiles with 1 to 4 secondary categories: 0.9 positions higher than profiles with 5+
- Profiles with all attributes filled (services, accessibility, payments): 0.7 positions higher
- Profiles with the 'Identifies as women-owned' or equivalent attribute set: small effect (0.2 positions) but visible in CTR
- Profiles with services marked up with descriptions and prices: 1.1 positions higher in the local finder
Review velocity correlated with map pack rank at r=0.41 in our cohort. Star rating, controlled for volume, came in at just r=0.18.
NAP, Citations, and Off-Profile Signals
Citations are the most over-discussed local SEO signal in the industry. Our cohort data agrees that consistency matters and disagrees that volume matters. Adding more citations beyond a clean baseline of the major aggregators (data axle, Foursquare, Localeze, Apple Maps) had near-zero correlation with rank movement.
- Profiles with NAP consistent across the four major aggregators: 1.4 positions higher than profiles with one inconsistency
- Adding 50 net-new citations per quarter: near-zero correlation with rank (r=0.06)
- Profiles with a city-specific landing page on the website: 0.9 positions higher
- Profiles with the GBP linked to a website on HTTPS with a 200 status and matching city in title tag: 0.7 positions higher
- Branded search volume growth correlation with rank: moderate positive (r=0.34)
What Did Not Move the Needle
We logged everything, including signals that did not correlate with rank movement. These are signals the industry talks about that did not hold up in the cohort. We are publishing them because pretending every variable matters is how budgets get wasted.
- Q&A volume on the GBP: no measurable rank correlation
- Posts published to the GBP weekly: weak correlation, mostly visible in CTR not rank
- Booking links added to the profile: small CTR lift, no rank effect
- Citation count beyond the four-aggregator baseline: near zero
- Profile age: weak correlation that disappears once review velocity is controlled for
Behavioural Signals From the SERP
Click-through and engagement signals from the local pack matter, but they matter inside narrow windows. Profiles with a higher CTR than their pack neighbours tended to climb. Profiles whose customers requested directions or called from the listing climbed faster than profiles with the same star rating but lower engagement.
- CTR delta of +20% vs pack average correlates with +1.2 positions over 60 days
- Direction requests per 100 impressions correlate with rank stability (r=0.29)
- Calls from the listing per 100 impressions correlate with rank (r=0.25)
- Average dwell time on the GBP profile expansion: not directly measurable, but profiles with high photo counts had longer engagement windows
What This Means for a 2026 Local Strategy
If you take five things into a local SEO plan from this dataset, take these. First, hold a steady review velocity (one to two new reviews per week minimum) instead of chasing a single quarterly push. Second, audit your primary category against the most specific match for the service you are actually selling. Third, upload owner-photos with location-correct EXIF every month. Fourth, fix NAP across the four major aggregators once and stop spending on long-tail citation services. Fifth, respond to 80%+ of reviews within 72 hours.
Everything else on the list is real but smaller. Spend your first 80% of effort on the five above. The remaining 20% on attributes, services markup, and city-page work pays off but only after the foundation is steady.
