Not showing in ChatGPT
The new search result is a sentence. Are you in it?
A growing slice of your customers now ask ChatGPT, Claude, or Perplexity who to hire — and get a paragraph naming two or three businesses. If yours isn't among them, you're absent from a results page you can't even see. This one is fixable, and early.
Same-day diagnosis. Flat quote before any fix.
What’s actually happening
AI assistants answer 'best roofing company near Mesa' from what they've learned and retrieved: your site's content, structured data, consistent business information across the web, reviews, and the citable authority of your pages. Absence usually isn't a verdict — it's illegibility: the models can't confidently parse who you are, where you serve, and why you'd be a safe recommendation.
This is generative engine optimization (GEO) territory, and it's unusually measurable now: you can ask the assistants yourself, today, and read the answer. We're deeper in this than most — we built AICiteBeam, a platform that tracks daily whether the AI models name and cite brands, against competitors. This page's diagnosis runs on the same discipline.
The usual causes, ranked
After twenty-seven years of these calls, the odds are well mapped. Start at the top.
Machine-illegible pages
No structured data, vague service descriptions, locations implied instead of stated — humans infer, models skip to a competitor they can parse.
Inconsistent entity signals
Name, address, services, and categories differing across your site, profiles, and directories — the model can't confidently merge you into one recommendable entity.
Nothing citable
Assistants favor sources they can cite: clear answers, real specifics, pages that resolve questions. Brochure fog gives them nothing to quote.
AI crawlers blocked
Robots rules or firewalls blocking GPTBot, ClaudeBot, and peers — sometimes deliberately, sometimes as a security plugin's default. Invisible by configuration.
What you can safely try first
Nothing below can make things worse — that’s the selection criterion. Anything riskier belongs in professional hands, on a backup.
- 1
Ask the assistants yourself
ChatGPT, Claude, Perplexity: 'best [your trade] in [your city]' and variants. Screenshot the answers. That's your baseline — free, immediate, and exactly what your customers see.
- 2
Check who they do name
The recommended competitors are your syllabus: what do their sites state plainly that yours implies? Usually it's services, areas, and proof, all machine-readable.
- 3
Check your robots.txt for AI crawler blocks
GPTBot, ClaudeBot, PerplexityBot in your disallow rules — or a security layer blocking them — means invisibility by choice. Two minutes to read.
Stop and call when…
- The models describe your business wrongly — correcting an entity takes structured, consistent signals, done properly
- Competitors dominate the answers in your own service area — that's a GEO program, not a tweak
- You want the change tracked over time rather than vibes-checked once
From there it’s our job: same-day look, flat quote, and the $229 flat repair covers most cases of exactly this.
Single Error Fix — buy it now, skip the hunt.
One error, hunted down and fixed — 500s, white screens, redirect loops, broken pages.
Covers one specific error or broken behavior on one site. Diagnosis, the fix, and a plain-English note on what happened. If we can't fix it, you get a full refund.
Questions we hear a lot.
Does AI visibility actually matter yet?
It's the fastest-growing referral channel we watch, and the answers concentrate hard: a sentence names two or three businesses where a results page named ten. Early movers in each market are capturing recommendation share while competitors haven't heard the acronym. The window is the opportunity.
How is GEO different from SEO?
It stands on SEO — the models learn from the same crawled web — then adds machine-legibility: structured data, entity consistency, citable answers, and AI-crawler access. Good SEO gets you into the model's knowledge; GEO work gets you into its recommendations. We do both, and they compound.
Can you actually measure this?
Yes — that's the unusual part of our bench: we built AICiteBeam to track exactly this, daily, across ChatGPT, Claude, Gemini, and Perplexity, with competitor share-of-voice. Your baseline, your trend, and whether the work is moving it — numbers, not folklore.
What does the fix cover?
The $229 diagnostic: your AI-visibility baseline across the major assistants, the illegibility audit (structured data, entity consistency, crawler access), and the prioritized fix list. The deeper GEO program — content, schema, tracking — is scoped from there, flat-quoted as always.
Related symptoms & help
Errors travel in packs. If this one visited, check its friends.
Get into the sentence.
Send the symptom, get a same-day look and a flat quote from the developer who's fixed this exact thing more times than either of us can count.