Methodology
How the numbers are made
Every AnswerGround number comes from Google Maps result lists collected at points around a business. This page describes the whole method: the data source, the grid, how your listing is recognised, and the formula behind each metric. If a number on your map surprises you, the reason is here.
Last updated
Where the ranking data comes from
Each point is a separate Google Maps search located at that point's coordinates. The results are collected for us by a third-party data provider, DataForSEO, as a mobile search that returns up to the first 20 listings. Manual scans run live; automatic rescans are queued with the provider and stored when the results come back.
Every scan is labelled with its source. Scans in Test Mode use generated sample data, cost no credits and are marked as sample data everywhere they appear. AnswerGround is not affiliated with Google.
How the grid is built
The business location is the centre. Around it, AnswerGround places concentric rings of points, evenly spaced out to the radius you choose. A scan has 15 to 90 points and a radius of 0.5 to 25 km. Outer rings are longer, so they get more points than inner rings. One scan can hold up to 10 keywords, and every keyword is searched at every point.
One credit is one point searched for one keyword. A scan of 40 points and 3 keywords therefore uses 120 credits, and the app shows that number before you start.
How your listing is recognised
When you add a business through the business search, its Google place ID is saved with it. At every point, the first listing with that place ID is your rank. Without a place ID, a listing counts as yours only when its name matches after normalising case, accents and punctuation, and it lies within 1 km of your saved location. That keeps another branch of the same chain from being counted as you.
The metrics
| Metric | Formula |
|---|---|
| Rank at a point | Your position in the result list at that point; “20+” when you are not in the first 20 |
| Share of local voice | Points where you rank 1 to 3, divided by all points |
| Top-10 coverage | Points where you rank 1 to 10, divided by all points |
| Average rank | Mean rank over all points, with “20+” counted as 21, so not appearing never helps |
| Competitor figures | The same formulas for every business that appeared, over the points that answered |
| Wins | Points where a competitor ranks above you, or appears where you do not |
Where you lose, and the likely reason
For every point where you are not in the top 3, AnswerGround looks at the business in position 1 and compares it with you on three things it can measure. It names the first one that applies:
- Closer: the leader is more than 0.4 km closer to that point than you are.
- Better rated: the leader's rating is at least 0.2 stars higher than yours.
- More reviews: the leader has more than 1.5 times your number of reviews.
If none applies, the point says the two look similar on distance, rating and reviews. Google weighs many more signals than these three, so this is a likely reason, not proof.
Points that fail
Sometimes the provider does not answer for a point in time. That point is marked as failed, left out of every metric, and its credit is returned automatically. If more than 25% of a keyword's points fail, the keyword is not stored at all and all its credits are returned.
Comparing scans over time
Compare puts two scans side by side point by point, only when they measure the same thing: the same business, keyword, point layout, radius and data source. A Test Mode scan is never compared with live data. Paid plans can repeat a scan automatically: weekly on Local, daily or weekly on Agency and Studio.
What we do not measure
There is no visibility score out of 100, no estimate of calls or revenue, and no promise of a ranking. Results can differ from what one person sees, because their device, account and history also play a part. For the general idea behind grid tracking, read geo-grid rank tracking, explained.