Stage 02 · Diagnose

Why AI recommends your competitors instead of you

Diagnose goes question by question and engine by engine. It shows where you lose, who wins instead, and why the AI preferred their pages: access, content, coverage, citations or perception.

Get my free reportNext: Prescribe
Gap analysis showing one critical gap, three high-impact gaps and five gaps in total, a per-engine breakdown of mention, share of voice and citation, and a crawler access panel showing which AI crawlers can reach the site.
Diagnose: where the brand is losing, engine by engine, and why. Demo data.
The question

What's wrong?

A low score tells you there is a problem. Diagnose tells you which problem. AI engines pass a brand over for a handful of reasons, and each one has a different fix.

  1. StarRise AI readiness overview for a demo brand: an AI Readiness Score of 51 out of 100, split into Brand Visibility 51 and Technical Readiness 50, with the brand named in 92% of answers, cited in 50%, and ranked second of six tracked brands.

    01 · Am I visible?

    Discover

    Your score, your rank and your results on every engine.

    Discover

  2. 02 · What's wrong?

    Diagnose

    Where you lose, to whom, and why.

  3. Weekly action plan showing the AI Readiness Score today at 51 and 77 projected when the plan ships, with the projected lift split across content coverage, content quality, citation authority and bot access.

    03 · How do I fix it?

    Prescribe

    A ranked plan with projected lift and hours.

    Prescribe
  4. Execution workspace with six tasks grouped by status (in progress, in review, to do and completed), each with a due date and its expected score lift.

    04 · Get it done.

    Execute

    Briefs, drafts and hands-on help.

    Execute
  5. Score tracking chart of the AI Readiness Score rising from 11 to 51 across six weekly runs, with changes against the prior run for brand visibility, mention rate, share of voice, citation quality and competitive rank, and a proof-of-value panel per AI engine.

    05 · Did it work?

    Measure

    What moved since your last run.

    Measure
What you see

Six views, one diagnosis

Each view answers one part of "why not us?" Together they point at the fixes worth making first.

Per-engine breakdown

Your mention rate, share of voice and citation rate on each engine, plus the competitor table filtered engine by engine. You see where one engine likes you and another ignores you.

Prompt gap map

Every buyer question, grouped by buyer stage, with your average rank or "Missing". It shows the share of each stage you win and which questions you hand to competitors.

Sentiment & threat

How positively each engine talks about you and each competitor, and which competitors threaten your position most.

Content quality factors

Your pages and your competitors' scored on what AI engines reward: statistics, quotations, cited sources and clear answers. The biggest gaps against the strongest competitor stand out.

Citation gaps

The exact URLs AI engines cite for your competitors and not for you, ranked by how much they matter. These are the sources AI trusts in your category.

Perception

Each engine is asked directly about you and each competitor. You see the strengths, weaknesses and terms it attaches to your name, in its own words.

Every buyer question, every answer

Prompt Explorer shows each buyer question by stage, whether you were named and cited, your position, and which brands and sources the answer named instead.

Prompt Explorer listing buyer questions grouped by buyer stage, showing for each question whether the brand was mentioned and cited, its position, and which brands and sources the AI answers named.
Discover: every buyer question, every answer, and who it named. Demo data.
The usual reasons

Four checks, in the order they matter

  • Can AI reach your site? A default "block AI bots" setting can hide a whole site. If crawlers can't read your pages, nothing else works. AI crawler guide
  • Is your content good enough to quote? AI engines favor pages with evidence. Opinions without numbers lose to pages that have them.
  • Do you answer the questions buyers ask? If you have no page for a buyer question, a competitor's page answers it.
  • Do the sources AI trusts mention you? Review sites, Reddit threads, articles and social posts. Where competitors are cited and you aren't, the gap shows.

Perception · example

How one engine describes "Your brand"

Illustration. You see this for every engine and every competitor.

  • Strengths: comprehensive, flexible, measurable outcomes
  • Weaknesses: limited public visibility, few published results
Which plan includes it

Diagnosis grows with your plan

Every plan shows how AI describes you. The full diagnosis of your website and social channels comes with Pro's weekly deep analysis.

PlanWhat Diagnose covers
FreeHow ChatGPT describes you, and your rank against one competitor.
StarterHow each engine describes you, with results on four engines.
ProWeekly deep analysis. AI diagnosis of your website and social channels, and competitor social media benchmarking.
White GloveEverything in Pro at a larger scope, with our GEO team acting on the findings.
See pricingNext: Prescribe, how do I fix it?
FAQ

Questions about Diagnose

Why does AI recommend smaller competitors over my brand?

Size doesn't decide it. AI engines name brands whose pages they can read, whose content answers the question with evidence, and whose names show up in sources they trust. A smaller competitor with a clear pricing page and a few Reddit mentions can beat a bigger brand with none.

What is the prompt gap map?

A grid of every buyer question you track, grouped by buyer stage. Each cell shows your average rank across engines, or "Missing" when the engines answered and didn't name you. Hover a cell to see the question. It shows which stages you win and which you give away.

Is AI saying something wrong about my business?

The perception view shows how each engine describes you today, in its own words. If an engine repeats a wrong or outdated claim, you see it and can fix the source: the pages and profiles the engines read.

What gets a page cited by AI engines?

What a careful editor wants: a clear definition up top, an answer paragraph that can be lifted as is, real numbers with a source, a named expert's quote, and a comparison table where options exist. Diagnose scores your pages on those factors against your competitors. More on the methodology page.

Does Reddit really matter?

For many categories, Reddit threads and review sites are among the sources engines cite most. Diagnose shows the exact threads and pages competitors are cited from, so you know where to take part as a practitioner.