Scoring Methodology

45 Local Search Ranking

We take a set of Google searches that this kind of business really gets in its area, and check where the business ranks for each one in Google's regular (organic) results. Higher positions earn more. A search where the business does not appear earns nothing, and it still counts in the average.

• Volume-weighted. Searches are averaged with a weight equal to the square root of their monthly search volume, so the searches your market runs most count more, without one search deciding the whole result. Where Google reports no volume for a search, we use the median of the others.
• Map pack. Up to 3 of the 45 points come from Google's local map results: the share of those searches where the business appears in the map pack, times 3.

35 Reviews

Two things from the business's Google listing: the star rating (about 16 of the 35 points) and the number of reviews (about 19). Each is read from a fixed table, so the same rating or count earns the same credit for every business. The tables show how much of that half is earned.

A business we can prove has no Google listing at all scores zero here, because there is nowhere for a customer to leave a review.

10 Competitive

We count the nearby competitors that have more Google reviews than the business does. The fewer there are, the more of the 10 points it earns. This is the one category that compares a business with others; everything else is scored against the fixed tables above.

10 Website

Google PageSpeed's mobile performance score for the business's website, from 0 to 100, counted in a straight line: a mobile score of 70 earns 7 of the 10 points.

AI search visibility Not part of the score

For RepAlmighty customers we measure whether ChatGPT and Gemini name the business when someone asks for the best in their area, and report it in the customer dashboard. It is measured and reported, but it carries 0 points and does not change the score. The free report explains AI search as a channel; it does not measure it.

Specialties found by referral

A few specialties are reached mostly by referral and reputation, not by searching, and their own searches carry very little volume. For those, the same four categories are weighted differently so the score measures what actually matters to that market. The categories, tables and rules above are unchanged; only the weights move.

• A category could not be checked. If a data provider fails or a check times out, the score is calculated from the categories that were measured, out of the points those categories are worth, and the report says how much was covered, for example "Based on 3 of 4 checks". Nothing is scored as zero because of our outage.
• The search check failed. Ranking is the largest category, so if it cannot be measured we do not show a score at all. We show what was measured and offer a retry.
• We could not confirm the Google listing. If we can neither find a Google Business Profile nor establish that there is none, the score is held at 59 or lower, the top of Needs work, so a report that did not look at the map, rating or reviews cannot read as Solid. Nothing is deducted; a lower earned score stays as it is.
• Customer accounts. A RepAlmighty account is re-measured by the platform, and a category that fails is carried forward from its last measurement for up to 21 days. Past that, the account keeps its last score instead of guessing.

Scores come from the same public data and the same formula for every business. When a business has a RepAlmighty account, the free report shows the same score as that account's dashboard, so the two match.

Only the providers we actually call are listed.

A free report is stored for 14 days. Running the same website again inside that window shows the same result; after it, a fresh report runs. If we change how the score is calculated, older stored reports are re-run instead of replayed. A customer's own numbers are re-measured by the platform on an ongoing schedule, so they do not depend on this window.

• We do not change a score based on the website platform a business uses.
• We do not grade on a curve. Apart from the Competitive category, every score comes from the fixed tables on this page.
• We do not adjust a score because a business competes with RepAlmighty.
• We do not edit scores by hand. A score is calculated by the scoring code; the only time the report shows a different number is the account dashboard score described above.

If you believe your score is inaccurate because of a data error, contact grader@repalmighty.com with your business name and the specific score in question. We will review the underlying data within 5 business days.

This methodology was last updated September 21, 2026. When the scoring formula changes, this page is updated with it.