How AI Perception works
AI does not evaluate your company only by what you say on your website. It looks at the larger picture of how your business appears online — your website, reviews, search results, citations, articles, competitors, and the way AI models describe your brand.
Think of this as your company's reputation through the eyes of AI.
You may describe your business as a premium service provider — but if AI only finds generic service pages, limited reviews, inconsistent citations, and no clear proof of results, it may describe you as a basic provider, or recommend a competitor with stronger trust signals.
What we measure
This report measures how AI understands and evaluates your brand as a whole — whether it can identify you, explain your value, find proof you can be trusted, and decide whether you're worth recommending.
How it actually works
AI Perception measures the quality of the mental model AI has of your business — not just whether it mentions you, but whether it can recognize you, describe you accurately, tell you apart from competitors, trust you, and ultimately recommend you. Visibility is being found; perception is everything that decides whether you get chosen.
A top AI model reads eight of your most important pages — chosen intelligently, because case studies, testimonials, and proof are often deep in the site, not on the homepage — and scores how well an AI could understand and vouch for you from that content. It's told to grade critically and honestly, not generously.
The real results from running buyer-style prompts through ChatGPT, Perplexity, Gemini, and Grok — how often you actually showed up, where you ranked, and whether you appeared on prompts written for your specific audiences.
The heart of it: a reasoning chain
Read it like a ladder AI climbs before it recommends you. Each rung depends on the one above it.
The first five rungs are the AI's judgment of your site — with concrete, evidence-bound fixes: never “add testimonials,” but “you claim X on this page yet Y is missing.” Fit and Recommended are counted, not guessed — they come straight from real prompt appearance rates and rankings. “Does AI recommend you” isn't an opinion; it's measured.
Walk the ladder top to bottom: recognize → understand → position → differentiate → trust → fit → recommend. Then find the biggest drop between two consecutive stages — that's the bottleneck breaking your momentum on the road from “found” to “chosen.”
One thing: your weakest rung, and the fix attached to it. You don't work all seven. Perception is the diagnosis; the Playbook is the prescription. Fix the drop-off, re-scan, and watch it move up the funnel.
Search used to send you ten blue links. AI sends one answer — either you're in it or you're invisible. There's no page two. SEO tells you where you rank; this tells you whether AI understands and trusts you enough to name you — before your competitors figure it out.
It scores how well AI understands and would recommend you, broken into the seven steps AI takes before it names you in an answer. You read it by finding the biggest drop-off — that's where you're falling out of AI answers. You fix that one thing, then re-scan. It matters because AI gives buyers one answer now, not a list — and this tells you whether you're in it, and why.
Prompt Tracking tells you if you showed up. AI Perception tells you why — and what AI would have to believe about you to recommend you next time.
What your gaps mean
Every pattern of scores points to a specific kind of fix.
The goal is not just for AI to know your company exists — it's for AI to understand, trust, differentiate, and confidently recommend your brand.