What is Google Maps rank tracking?
Google Maps rank tracking is measuring the position of a business in Google Maps results for a keyword, from a defined location, repeated over time. Because Google weighs the distance between the searcher and the business, the position changes with the search location, so a useful tracker checks many locations around the business (a geo grid) rather than one.
That last point is the whole argument of this post, and it is usually made with a diagram. We wanted to make it with data instead: how much does a Maps position actually vary across a few kilometres, how often does the classic single check mislead, and what does that imply for how often and from where you should track?
If you only want to see your own spread first, the free Google Maps rank checker runs a 3×3 grid without an account. The mechanism behind the variation is covered in how the proximity factor works; this post is about measurement.
The sample and the method
Every figure below comes from one query set run on Geogrid's production database on 19 October 2026. Only aggregates were read: no business names, keywords or locations appear in this post.
What is in the sample. Completed Google Maps scans created between 16 September and 18 October 2026, not deleted. Scans of Google web results are excluded. We start on 16 September because that is when Geogrid began choosing a location Google actually recognises for each grid point; earlier scans are compared separately further down. Grid points that fall on water are stored as unranked and would read as false losses, so they are removed. 222 of the 261 scans are 5×5 grids, 37 are 7×7 and 2 are 13×13; 182 use a 5 km radius, which puts neighbouring points 2.5 km apart on a 5×5.
What a rank means here. Each point records the business's position in the first page of Google Maps results for that keyword at that location, which is 10 results. A point is either a position from 1 to 10 or "not in the top 10". When a business is in the top 10 nowhere on the grid, a second probe reads page two at 9 points of the grid (corners, edge midpoints, centre) and stores the best position found.
How we group scans. "Visible" means the business is in the top 10 at one or more land points. "Patchy" means it is in the top 10 at some points and outside it at others. All percentages are counts of scans, not of points, unless stated.
Finding 1: half of visible businesses have no single Maps rank
Of the 261 scans, 117 found the business in the top 10 at least once. Those 117 split almost exactly in two: 59 were in the top 10 at every land point of the grid, and 58 were in the top 10 at some points and outside it at others.
Reaching the top of the map does not protect a business from disappearing a few kilometres away. Of the 53 businesses that held #1 at one or more points, 19 were outside the top 10 at other points of the same scan. Of the 79 that reached the top 3, only 21 (27%) held a top-3 position at every point, and the median business in that group was top 3 at 57% of its points.
For a tracker, this is the difference between reporting "you rank #2" and reporting "you rank #2 near your address and are absent from the first page in the east of the city". Both statements come from real measurements. Only the second one tells the client where the work is.
Finding 2: a check from the centre point misleads about one time in four
Most rank checkers, and most manual checks, look from one place: the business address or the city centre. We compared the position at the centre of each grid with the rest of the same grid.
We placed each point in one of three brackets: top 3, positions 4-10, or not in the top 10. In 28 of the 116 visible scans that had a centre point (24%), the centre's bracket described less than half of the grid. In the median scan it described 79% of the points, which is good but not something to build a monthly report on.
The centre is also the friendliest point. Among visible scans, the business was in the top 10 at 85% of centre points but at 74% of points on the outer edge of the grid (top 3: 44% against 39%). And the error runs both ways: of the 159 scans where the business was outside the top 10 at the centre, 17 found it in the top 10 elsewhere on the grid, 7 of them in the top 3. A one-point check would have told those 17 businesses they were invisible.
Note that the grid centre is wherever the user placed it, which is not always the business's own address. That is realistic, since people also check from wherever they happen to be, but it means this is not a pure distance measurement. For checking a specific neighbourhood or suburb, see how to check Google Maps rankings by location.
Finding 3: the #1 spot changes hands across one map
If positions vary across a grid for one business, they vary for its competitors too. We counted how many different businesses hold the #1 position across a single grid.
This uses the 78 scans in the sample where the #1 result was recorded at every land point (the competitor list is not stored for every older point). In under a third of them a single business owned #1 across the whole area. In 42% the top position was split between three or more businesses, and one grid had 19 different leaders.
Practically, "who is our main competitor on Maps?" rarely has one answer for a whole service area. The competitor that matters depends on the neighbourhood, which is another reason a single-point tracker under-reports the picture.
Finding 4: invisible on page one usually means far from it
144 of the 261 scans (55%) found the business in the top 10 at no land point at all. The page-two probe then found it at positions 11-20 in 28 of those 144 scans (19%), with a median best position of 15.
So when a grid comes back entirely outside the top 10, about one business in five is near the edge of page one somewhere and four in five are further away, beyond position 20 at all 9 probe points. Those are different problems. The first is a ranking gap that normal optimisation can close; the second is more often a relevance problem (category, keyword fit) or a listing issue, which we walk through in why a business is not showing on Google Maps.
The share varies by sector, with the caveat that our industry groups are small. The only groups with 20 or more scans were home services (10 of 42 scans visible, 24%), beauty and wellness (11 of 23, 48%) and healthcare (8 of 20, 40%), against 45% for the whole sample. Home services, a sector with many service-area businesses that have no storefront, had the lowest visibility in our data.
This number is also skewed by who runs scans. Most of the sample comes from free accounts, and people tend to scan when they suspect a problem, so 55% is not an estimate of how invisible local businesses are in general.
Finding 5: bad location handling invents drop-offs
The same queries on the 320 completed scans from before 16 September give a very different picture: 77% of visible businesses look patchy (119 of 154), against 50% in the current sample, and 77% of top-3 businesses also show points outside the top 10 (86 of 111), against 37% now.
Part of the gap is simply a different set of users and keywords. Most of it is how each point's location reached Google. Geogrid gets Maps results through a search API that takes coordinates plus a place name, and we found that Google only honours the place name when it recognises it; otherwise it silently falls back to the country or even to another country. Before the fix, 47% of points in a 20-scan test returned results that were actually local; after it, 95%. A point that silently searched somewhere else showed the business as absent, which looks exactly like a real drop-off on a heatmap.
The lesson applies to any tool, not just ours: a Maps rank tracker is only as good as its proof that each point's results come from that point. When you compare local rank trackers, ask how they set the search location and whether they check that the results returned are near the point. A red patch on a grid is worth acting on only if the tool can show that the competitors it found there are actually nearby.
How to track Google Maps rankings, based on this data
Turning the findings into a tracking setup:
- 01Track a grid, not a point. Half of visible businesses in our data had both top-10 and non-top-10 points on the same map. A single position cannot represent that, whichever point you pick.
- 02Report the share of the grid in the top 3 and top 10, not an average rank. An average of #2 and "not in the top 10" is not a position anyone sees. Shares survive patchiness and are easy to compare from one scan to the next.
- 03Size the grid to the real service area. Most scans in our sample used a 5 km radius 5×5 grid, 2.5 km between points, and that spacing was already enough to put top-3 and absent points on the same map. Wider areas need more points to keep the spacing comparable.
- 04Keep everything fixed between scans. Same centre, radius, grid size and keyword wording. Otherwise a change in the map is a change in the method.
- 05When the whole grid is outside the top 10, look at page two before rewriting the listing. Positions 11-20 call for different work than being absent from the top 20.
- 06Check one recurring scan per important keyword rather than many one-offs. Our sample has too few repeat scans to say how fast grids move, so pick a cadence that matches how often you change the listing: weekly during active work, monthly for maintenance.
Limits of this data
This is an observational sample of scans people chose to run, not a random sample of businesses. It is dominated by free accounts and small grids, spread across 26 countries, and concentrated in the five weeks after a methodology change. Each scan is a snapshot: we cannot say from it whether a patchy grid is stable or noisy over time, because only a handful of scans in the window repeat the same business, keyword and grid.
Ranks come from Google Maps results obtained through a search API at each grid point, not from a real phone at that spot, and they stop at position 10 on the grid (20 for the page-two probe). The centre comparisons use the grid centre chosen by the user. We will rerun these queries as the sample grows and update the figures with the date and sample size.
See your own spread.
Run a grid on your business with 200 free credits, no card needed. You see where on the map you rank and where you drop out, and Geogrid checks page two if the grid comes back empty. Then keep one weekly rescan on the free plan.