We analyse content, structure, menu and information the way AI systems read it
Categories, reviews, photos and updates AI uses to understand your business
Nearby hotels, landmarks and businesses AI weighs when recommending you
Hundreds of real questions asked across ChatGPT, Gemini, Claude and Perplexity
Which competitors AI recommends instead of you, and why
Reviews, mentions and other authority sources AI relies on
What each metric means
Tap a metric to see how it works.
AI Visibility Score
A 0–100 score combining mention frequency, intent coverage, model consistency, and website signals, one number to track your visibility over time.
Mention Rate
The share of tested AI responses that mentioned your restaurant. A rate of 42% means you appeared in 42% of relevant answers.
AI Model Coverage
Which AI assistants (ChatGPT, Claude, Gemini, Perplexity) recognise and recommend you. Consistency across models matters more than strength in just one.
Competitive Position
How you compare to the restaurants AI recommended instead of yours, for the same customer intent. This comparison set changes depending on what was asked.
Intent Coverage
How many distinct customer situations (romantic dinner, family lunch, business meeting) your restaurant is visible for, not just keywords.
Strong Intents
The customer intents where AI consistently recommends your restaurant, worth protecting.
Opportunity Areas
Customer intents where AI rarely recommends your restaurant, the highest-impact places to improve.
Visibility Trend
Compares your latest scan with previous ones to show whether visibility is moving up or down.
Expected Impact
An estimate, based on patterns from similar restaurants, of how much a recommendation could help, not a guarantee for your case.
Implementation Effort
An estimate of how much work a recommendation requires, used alongside Expected Impact to prioritise.
Confidence
How consistently a recommendation's pattern has been observed across restaurants, high confidence means well-established, lower means emerging.
Five steps, every scan
Frequently asked questions
Why isn't my score 100?
No restaurant is recommended for every possible customer intent, by every AI model, every time, a certain amount of variation is normal, even for well-established restaurants. A very high score is achievable, but 100 would mean perfect, universal recommendation, which isn't a realistic or meaningful target. Finded focuses on measurable progress, not a perfect number.
Why do competitors change?
The restaurants AI recommends instead of yours depend entirely on what's being asked. A different customer intent surfaces a different set of alternatives, so your competitive set shifts from one evaluation to the next. This is expected, it's exactly why we track competitive position per intent, rather than as a single fixed ranking.
Can AI recommendations change?
Yes. AI models are updated regularly, and their responses can shift as a result, independent of anything you've changed on your end. This is why Finded monitors visibility continuously through repeat scans, rather than treating a single measurement as final.
Can Finded guarantee recommendations?
No. Finded measures and reports what AI models currently return, and provides evidence-based recommendations to improve your visibility. We don't control AI models and we don't attempt to manipulate their outputs, we can't guarantee specific outcomes, only give you an accurate, evidence-based view of where you stand and what tends to help.
How often should I scan?
Monthly is enough to track meaningful change for most restaurants. If you've just implemented recommendations, a rescan lets you verify whether they had a measurable effect, rather than assuming they worked.
Why not just ask ChatGPT?
A single question gives you one data point, from one model, at one moment. Finded evaluates a structured set of customer intents across multiple AI models on a recurring basis, so you get a measured, repeatable picture, including trend and competitor data, instead of a one-off answer.
