How Localist ranks businesses

A short, honest explanation of the formula behind every directory page — version 2 of our methodology. The full white paper, with proofs, edge cases and a reference implementation, is available for download.

Localist Ranking Methodology — v2 (white paper) PDF · proofs, edge cases & a reference implementation

Download PDF

Three inputs, one score

Every business on Localist is scored from the same public information on its Google Business Profile:

Rating — the 1.0–5.0 star average shown on Google.
Review count — how many customers have actually left a review.
Recency — the timestamps of the five most recent reviews.

Rating alone isn't enough: a 5.0 from one review proves almost nothing, and a business that stopped earning reviews three years ago may not be the business it once was. So we combine all three into a single Localist score.

The formula

Every page compares businesses only against their local market — the same city, the same vertical. Three numbers describe that market, and we publish them on every page:

C — the average star rating of all businesses on the page
m — the typical review count for the market (clamped between 10 and 50)
g — the typical age, in months, of a business's fifth-newest review

Then each business is scored in two steps:

Step 1 (evidence): v1 = (n·R + m·C) / (n + m) Step 2 (freshness): F = 0.5 ^ (max(0, a − g) / 24) Localist score: v2 = C + F · (v1 − C)

where R is the business's rating, n its review count, and a the age in months of its fifth-newest review.

Plain English: every business starts at the local average. Each verified review pulls its score toward its own rating — the more reviews, the stronger the pull. And if a business has stopped receiving reviews, that pull fades with a 24-month half-life, letting the score drift back toward "average" — because old evidence is weak evidence, in either direction.

Worked examples

Same market (local average C = 4.5, m = 50, g = 3):

Rating	Reviews	Last 5 reviews	Localist score	What it means
4.9	87	recent	4.75	Excellent and actively proven
4.9	20	recent	4.61	Excellent, still building evidence
4.8	400	3 years old	4.60	Great record — but a stale one
3.2	500	recent	3.32	Volume can't outrank quality
5.0	3	recent	not ranked	Too little data to make a claim

Notice the second and third rows: twenty recent reviews at 4.9 now edge out four hundred stale ones at 4.8. Evidence counts; current evidence counts most.

How a directory page is built

Eligibility A business needs at least 5 reviews to be ranked. Below that, we show its stars and count with an "insufficient data" label instead of a score.

Merit list Businesses scoring at or above the local average (v2 ≥ C) form the headline ranking: they are demonstrably above their market's norm.

Also notable Remaining eligible businesses, in score order.

Ties are broken by review count, then by most recent review. Rankings refresh monthly from the latest data we hold, and every page shows when.¹

Paid placements never change organic positions. Featured Partner slots, where offered, live in a strictly separate block. That's a hard guarantee, not a setting.

Flags, not fudges

Some things the score deliberately does not absorb. Instead, we show them as visible labels:

Trending lower the last five reviews average at least 1.5 stars below the lifetime rating (only shown for businesses with 50+ reviews).

Dormancy note review activity has been quiet for well beyond the local norm.

Held for review suspicious patterns (a sudden burst of reviews after long silence, near-duplicate wording) pause a listing for manual review.

A hidden "trust adjustment" inside the formula would make our rankings impossible for you to verify. Visible flags keep the score recomputable and still tell visitors what the number can't.

Check our math

Take any business on any page: its stars R and review count n from Google, plus the C, m and g printed on the page. Run the three lines above. You'll get our number to two decimal places — and if you don't, we want to know: ai@konigle.com.

Didn't make the list?

Not ingested yet? Use the Get listed button on the page — you'll see your projected score before you submit.
Score below the local average? That almost always means not enough recent reviews, not bad ones. Ask happy customers to leave a Google review — steadily, not in a burst. Both your evidence and your freshness improve.
Strong credentials elsewhere? Attach them to your listing request. Staff can fast-track inclusion on off-platform evidence — but nothing ever adjusts a score.

What we will not do: weight self-reported testimonials, paid reviews, or off-platform claims into the score. The formula only looks at what Google has publicly verified. That's the whole point.

Questions or disputes: ai@konigle.com.

1. How fresh is the data behind a score? Fetching Google Business Profile data at scale is expensive, so we refresh in tiers: claimed businesses are re-fetched very frequently, and every business — claimed or not — is re-fetched at least once a year, so no listing is ever ranked on data older than 12 months. If your business is unclaimed and your listing looks out of date, claiming it (via the Get listed button) is the fastest way to get onto the frequent-refresh tier.