Which local SEO KPIs to track, and which to ignore
Every metric in the same format: what it measures, how you calculate it, how you read it. Plus the two metrics that suggest a performance which does not exist.
Three metrics are enough to run local visibility: top 3 share of voice, the weighted visibility score and the visibility radius. All three come out of the same reading — one rank per keyword × location pair — and read without interpretation. Two very common metrics, on the other hand, describe local performance badly: average position and impressions.
Top 3 share of voice
What it measures. The proportion of situations in which your profile appears in the top three local pack results. It is the reference metric, because the local pack only shows three businesses: fourth place is commercially close to tenth.
Calculation
Top 3 share of voice = (measurements ranked 1 to 3 ÷ total measurements) × 100
Example — 42 top 3 appearances across 120 keyword × location pairs gives 35% share of voice.
How to read it. It is a percentage, so it is comparable between locations of different sizes and between periods, even when the number of measurements changes. That is its main quality: a network can rank all thirty of its sites on this single column. A change of less than three points month to month is generally noise.
Weighted visibility score
What it measures. The same thing as share of voice, but taking the exact rank into account rather than a binary threshold. It addresses a genuine weakness of the previous metric: moving from 4th to 2nd changes nothing for a share of voice that already counts the appearance, even though it is a major improvement.
Calculation
Weighted score = Σ (rank weight × presence) ÷ number of measurements
With a decreasing weighting, for instance rank 1 = 1.0 · rank 2 = 0.8 · rank 3 = 0.6 · ranks 4 to 10 = 0.2 · outside top 10 = 0. A profile ranked 1st on 10 measurements and 5th on 10 others scores (10 × 1.0 + 10 × 0.2) ÷ 20 = 0.60.
How to read it. The absolute number means nothing on its own; what counts is its movement, and the comparison between locations on identical weighting. Decide the weighting once and leave it alone: changing it rewrites the entire history.
Why not just average the ranks
An average of ranks treats the gap between 3rd and 4th as equal to the gap between 15th and 16th. But the first drops you out of the local pack, and the second changes nothing for anyone. A decreasing weighting makes that step visible; an average smooths it away.
Visibility radius
What it measures. How far out the profile is still visible. It is the most immediately meaningful metric for a business owner, and almost nobody calculates it.
Calculation
Visibility radius = the maximum (or median) distance at which the profile still ranks in the top 3 — or the top 10
Example — a location ranked 1st a quarter of a mile from its address, still in the top 3 at 1.5 miles, and outside the top 10 at 2.5 miles: its top 3 radius is 1.5 miles.
How to read it. Set against the real catchment area, it tells you immediately whether the problem is a visibility problem or a market problem. A 1.5-mile top 3 radius for a shop whose customers come from within a mile: visibility is not the issue. The same radius for a tradesperson working across fifteen miles: it is the whole issue.
Publish it at both thresholds, top 3 and top 10: the gap between them describes how fast visibility collapses as you move away, which no other metric shows.
Profile actions: calls, directions, clicks
These three numbers come from the Business Profile's own statistics, not from a rank tracking tool. They measure what visibility produces, and they are the essential bridge between a rankings report and a conversation with a board.
Track them as movement, never as absolute values: the definitions and scope of these statistics change as Google evolves, which makes long-run comparisons hazardous. And set them alongside rankings without confusing the two: you can gain three places and get no extra calls, because position was not the limiting factor.
Discovery searches and branded searches
A branded search — someone typing your name — does not measure your local SEO: it measures your brand awareness, usually built elsewhere. A discovery search — someone typing a trade or a service — is what local work moves.
Separating the two in the report avoids the most common illusion: a rise in the total driven entirely by branded searches while discovery flatlines. That is especially true after an advertising campaign, which inflates branded volume without changing anything else.
One figure not to repeat
A split of local searches into "72% discovery, 22% direct, 6% branded" circulates widely. On checking, that split does not appear in the statistics BrightLocal publishes today. Track the split on your own profile statistics: it varies enormously by trade, and a market average does not describe you.
The misleading metrics
Two widely used metrics describe local performance badly, and it is better to remove them from a dashboard than to explain every month why they mean nothing.
Avantages
- Top 3 share of voice, by named location and by keyword
- Weighted visibility score, on a fixed weighting
- Visibility radius, at both the top 3 and top 10 thresholds
- Calls, direction requests and website clicks, as movement
- Number of competitors ahead of you, at each location
Inconvénients
- Average position across all surfaces: it blends organic, local pack and Maps
- Search Console average position: it aggregates thousands of queries and places
- Impressions: they rise when you rank badly for a lot of queries
- Number of keywords "in the top 10": inflatable at will by adding variants
- Average rating on its own, without review count or freshness
Search Console's average position deserves an extra word, being the most quoted. It covers organic results, aggregates every query and every place, and moves mainly when the composition of that blend changes. A rise can mean you have stopped appearing at all for the queries where you ranked badly — which is a deterioration.
Building a dashboard readable in 30 seconds
A report is judged on what a director takes from it in half a minute. Four elements are enough, in this order.
The four elements, in order
- One number at the top: this month's top 3 share of voice, and the change on last month in points.
- A three-series chart: your share of voice, the average of the competitors you track, and the local leader. It is the only chart that shows whether you are progressing or the market is.
- A table by named location: this month's rank, last month's, and the gap. This is what you look at when you want to act.
- Three lines of commentary: what was done, what moved, what is planned. Without them, the report is an archive.
What must not be in it: any metric you could not explain in one sentence, and any number nobody will ever act on.
These metrics, calculated for you every month
Top 3 share of voice, movement by location, comparison against the competitors actually shown: Pinperf records your rankings from the locations you name and returns a dashboard you can put in front of a client or a board as it is.
Sources
- Google Business Profile Help — Improve your local ranking on Google.
- BrightLocal — Local SEO Statistics: market statistics on local search.