Grid size and spacing are two separate choices. Size is how many points you check: 5×5 is 25, 7×7 is 49. Spacing is the distance between neighbouring points, and spacing × (n − 1) is the width the grid covers. Pick the spacing from where your customers come from, then pick the smallest size that fills that area.
Size and spacing are two different settings
A 7×7 grid at 300 m spacing is 1.8 km across. A 7×7 grid at 2 km spacing is 12 km across. Same 49 points, same cost, two different questions. The small one asks “where in this neighbourhood do I win?”. The large one asks “in which towns do I appear at all?”.
Most mistakes with grids come from changing the size when the spacing was the problem. Going from 7×7 to 9×9 at the same spacing adds a ring of points around the edge and makes the grid a little wider. Doubling the spacing makes it twice as wide with no new points. The arithmetic is the same for every grid:
Match the radius to where customers come from
The radius is a fact about your customers, not about Google. Ask where the people who pay you actually set off from, and make the grid reach a little past that.
- A city-centre café: people walk. 1 km across is plenty, so a 7×7 at 150 m to 200 m.
- A dentist, a vet, a hairdresser: people drive 10 minutes. 3 km to 5 km across, so a 7×7 at 500 m to 800 m.
- A plumber or an electrician: you drive to them. The grid is your drive radius, often 15 km to 20 km across, so a 9×9 at 2 km or more.
- A restaurant people travel for: wider than it looks. Scan 5 km across before deciding.
The grid should reach the streets where you stop winning. If every point is green, the grid is too tight and tells you nothing you did not know. If the outer two rings are all 20+, it is too wide and half the points are spent on places no customer comes from.
Which grid size to use
With the spacing set, the size follows: enough points to fill the area at that spacing, and no more. These are the sizes most people settle on.
| Size | Points | Good for |
|---|---|---|
| 3×3 | 9 | A one-block check. Too coarse to read a trend. |
| 5×5 | 25 | A quick look at a storefront. The largest scan on Track Free. |
| 7×7 | 49 | The everyday grid for a shop, clinic or restaurant. The grid the free audit runs. |
| 9×9 | 81 | A town, or a business whose customers drive further than 10 minutes. |
| 11×11 to 13×13 | 121 to 169 | A service-area business, or a town with its suburbs, at 1 km to 2 km spacing. |
| 15×15 to 21×21 | 225 to 441 | A metro-wide brand or a multi-location report. Not for a single storefront. |
More points do not mean more precision. At the same spacing they mean more coverage; in the same area they mean finer detail. A 13×13 at 250 m sees every street in a 3 km patch, which is useful for a dense centre with rivals on every corner and wasteful anywhere else.
Square, circle or hexagon
A square grid spends a lot of its points in the corners. A corner point sits about 1.4 times further from the business than the middle of an edge, and in most towns that is a field, a motorway or a neighbourhood nobody drives from. A circle drops those points: a 7×7 circle is 29 points instead of 49, and the map still covers the streets customers come from. A hexagon trims the corners too and spaces its points evenly in every direction.
Choose the square when the corners matter: a service area that is a rectangle of towns, or a motorway junction in one corner that brings customers. Choose the circle when the business pulls evenly, which is most of them. For a service-area business ranking from a hidden address, the circle matches the shape of the reach.
Keep the same grid from scan to scan
A grid is a fixed set of searches. Change the spacing, the size, the centre or the shape and you have a different set of searches, and the numbers from the two scans measure different things. Average rank can move by two places with no change in any rank, simply because the far points were added or taken away. Share of local voice does the same.
So set the grid once, and hold these still from scan to scan: the centre, the spacing, the size, the shape, the keyword wording and the engine. Scan at roughly the same time of day. If you must change the grid, note the date and treat everything after it as a new series. How often to check local rankings covers the interval; ARP and ATRP covers what the averages do when points change.
RankShark Track keeps the grid fixed for each tracker, from 3×3 to 21×21 as a square, circle or hexagon, runs it on demand or on a weekly or monthly schedule, and keeps 25 months of history, so the scans line up without you having to remember the settings.
What a grid costs in points
A point is one search from one spot. A 7×7 square scan costs 49 points, a 7×7 circle 29, a 5×5 square 25. Divide a plan’s monthly points by the points in your scan and you have the number of scans a month.
Track Free gives 150 points a month, with scans of up to 25 points. A 5×5 grid for one keyword is 25 points, so 150 ÷ 25 = 6 scans a month: 2 keywords, 3 scans each, on demand. Enough to see where a business stands and whether a change moved the map.
Starter, the lowest paid plan at $29 a month, gives 10,000 points, with scans of up to 81 points, so a 9×9 square or a 7×7 circle fits. On a 7×7 square that is 10,000 ÷ 49 = 204 scans a month. Ten businesses with two keywords each, scanned weekly, is about 87 scans, or roughly 4,300 points, with room for a wider grid on the two that need it. On a 29-point circle the same plan runs 344 scans.
The pattern holds up the range: a bigger plan buys either more businesses at 7×7 or the same businesses at 13×13. The Track plans list the monthly points and the largest scan each one allows.
The next step: work out where your customers start from, set the spacing so the grid reaches just past that edge, pick the smallest size that fills it, and run the same grid again next month. The free audit runs a 7×7 if you want to see one before you decide.