why chatgpt doesn't recommend your business (and what fixes it).

elisabeth hitz · july 20, 2026 · 6 min read

a founder watches signups drop, asks chatgpt the question their buyers ask, and finds competitors listed as the default answers. that moment is arriving for every business whose customers research before buying.

chatgpt doesn't recommend your business because AI engines can only recommend what they can read, understand, and cite. if crawlers are blocked, your pages bury the answer, nothing on your site is citable, and no structured data says what you are, the engines fill the answer with competitors who fixed those things.

the shift already happened

in G2's 2025 buyer research, generative AI chatbots became the single most influential source for B2B shortlists, ahead of review sites and vendor websites. forrester's research finds a majority of buyers now use AI to compare vendors before they ever talk to a human. the answer box is the new first meeting.

the six checkable causes

checkthe question it answers
crawler accesscan GPTBot, ClaudeBot, PerplexityBot read you at all?
structured datadoes schema tell engines what you are and offer?
answer-first structurecan an engine lift a clean answer from your page?
citable authoritydo you publish stats, sources, and dated claims worth quoting?
comparison and faq surfacesdo pages exist where engines look for recommendations?
machine-readable summarydoes llms.txt hand engines your story directly?

how to check yours in about a minute

i built a free scan that runs those six checks against any live site and grades each engine's access: it is on the ai visibility page, and the result renders on the page, no waiting. the full checklist version, if you want to fix it yourself, is in the 2026 ai visibility checklist.

takeaway

being absent from AI answers is not a mystery or a penalty. it is a list of six fixable gaps, and the businesses fixing them now are inheriting the buyers who stopped clicking ten blue links.

see where you stand

run the free ai visibility scan.

six real checks against your live site, graded engine by engine, results on the page. the audit turns your gaps into a ranked fix plan against your actual competitors.

run the free scan →

written by elisabeth hitz, certified in anthropic's ai fluency program (framework & foundations, and ai capabilities & limitations), plus claude 101 and claude cowork.

sources: G2 2025 buyer behavior data via omnibound.ai; forrester buyer research via 6sense. related: how chatgpt decides what to recommend · the assistant is the new homepage

written by elisabeth hitz, certified in anthropic's ai fluency program (framework & foundations, and ai capabilities & limitations), plus claude 101 and claude cowork. framework: "AI Fluency: Frameworks and Foundations," a free course by Rick Dakan and Joseph Feller with Anthropic, released under CC BY-NC-SA 4.0 (anthropic.com/ai-fluency). the 4Ds, the three interaction modes, and the definitions are theirs. the business application, the "one you skip" read, and the setup advice are mine, from setting AI up on real businesses. adoption stat: IBM 2026 Global CEO Study (access vs. active-use gap).