The 7 Metrics Behind Your AI Visibility Score

Do AIs mention you at all?

Out of everyone who asks an AI “best plumber in Kwun Tong,” how many times does your name actually come up? Recall is the fraction of AI answers that name your business.

Recall is weighted highest because it's the foundation — you can't rank well, be described consistently, or be factually correct in an answer that never mentions you in the first place. Being invisible beats every other problem.

Example: if we sample 10 AI-generated recommendations for your category and city, and your name appears in 3 of them, your Recall is 30%.

How to improve it: fresh reviews, a complete and consistent Google Business Profile, and being listed on the directories AI engines actually pull from.

How consistently does each AI engine mention you?

Breadth is the average recall across each individual AI engine. If Gemini mentions you in 60% of its answers and Perplexity mentions you in 20%, your Breadth is the mean of those per-engine rates. It measures how deeply you're embedded in each engine's understanding — not just whether one engine found you once.

Different engines pull from different sources. A high overall recall driven by just one engine is fragile — if that engine changes its algorithm, you disappear. Breadth rewards businesses that are consistently visible across all engines, not just lucky on one.

Example: if we test 5 engines and each one mentions you in an average of 40% of its responses, your Breadth is 40%.

How to improve it: structured data (schema markup) helps web-retrieval engines find you live; a long-standing, consistent web presence helps trained engines "remember" you from their training data. Each engine has its own index — fix them one at a time using Visora's Engine Deep-Dive.

Do the AIs agree on who you are?

When two different AI answers both mention your business, do they describe the same thing — or contradict each other? Convergence measures how consistently AI engines describe you.

Imagine two friends describing you to a stranger. If both say the same accurate things, that builds trust. If one says you're a 5-star restaurant and the other says you're a nail salon, that's confusing — and confusion kills recommendations even when you technically got mentioned.

Convergence needs at least 2 real mentions to compare — with only 1 mention, there's nothing to compare against, so this metric shows as “not yet available” rather than penalizing you.

How to improve it: keep your category, services, and description consistent everywhere online. Inconsistent listings (different business names, categories, or descriptions across sites) are the most common cause of low Convergence.

Do the AIs get your facts right?

Out of what an AI says about you — your hours, services, location, credentials — how much of it is actually true? Factuality checks AI responses against the facts you've confirmed.

AI hallucination is real — engines can confidently state things that are simply wrong. An outdated address, a service you dropped years ago, hours that no longer apply. Factuality is the metric that catches this specific failure mode.

Example: if your business has 5 confirmed facts on file, and AI responses correctly reflect 3 of them on average, Factuality is 60%.

How to improve it: keep your Google Business Profile, website, and directory listings synced and current. Correct outdated information anywhere an AI might be pulling from.

When you're mentioned, are you #1 — or #5?

Being mentioned is step one. Being mentioned first is what actually drives the click or the call. Rank measures your position relative to competitors when an AI names several options.

Think about a list of 5 recommendations. If an AI says "here are 5 good options" and you're #5, most people never read that far down. Position matters almost as much as being mentioned at all.

1st place scores 1.0, 2nd scores 0.8, 3rd scores 0.65 — the score drops fast, because attention drops fast the further down a list you go.

How to improve it: Rank is mostly a downstream effect of the other four metrics — stronger Recall, wider Reach, cleaner facts, and more consistent descriptions all push you higher up the implicit ranking an AI builds when comparing options.

How much does the AI say about you?

When an AI mentions your business, does it write a full paragraph with details — or just drop your name in a list? Depth measures the richness of responses that mention you, based on how much text the AI devoted to your business (capped at 500 characters).

A brief name-drop is fragile visibility. A detailed, informative mention — where the AI explains what you do, where you are, why you're good — is what actually influences a customer's decision. Depth rewards businesses that give AI engines enough structured information to write about them substantively.

Example: if an AI writes a 500+ character description about your business, that response scores 1.0 for depth. A 100-character mention scores 0.2.

How to improve it: complete your Google Business Profile with a detailed description, add schema markup with services and offerings, and create content that gives AI engines rich information to draw from.

Across how many different questions does the AI mention you?

Customers don't all ask the same question. Some ask 'best plumber near me', others ask 'who should I call for a burst pipe', or 'recommended plumber in Kwun Tong'. Keyword Reach measures the fraction of different search queries where your business appears in the AI's answer.

Being mentioned for one query but invisible for all others is fragile. Keyword Reach rewards businesses with broad relevance — the more questions where AI engines think of you, the more customers you capture across the full spectrum of how people actually search.

Example: if we test 10 different search queries and your business appears in answers to 4 of them, your Keyword Reach is 40%.

How to improve it: create content that addresses different ways customers search for your services — not just "best X" but "how to", "who should I call", "X near me", "X reviews". The more query patterns you cover, the higher your Keyword Reach.