Ask most agencies where a firm ranks in the map pack and you get one number. That number is close to meaningless, because the map pack is recomputed for the searcher’s location. A single reported rank is really the rank at one point — usually the office, which is the most flattering point on the map.
We track on a grid of points spread across the market instead. The result is a coverage picture: strong near the office, fading with distance, with specific suburbs where you do not appear at all. It is a less comfortable report and a far more useful one, because the gaps are the actionable part.
An average rank is the arithmetic mean of the places you win and the places you are invisible. It is a way of not seeing the second group.
It also makes expansion decisions concrete. When the grid shows a suburb producing searches you never surface for, that is a documented argument for a second office — or for conceding it and spending the money on paid search instead, which has no proximity constraint.