Join the waitlist

Leave your email and we’ll let you know when Scaplet is ready for you.

We’ll only email you about getting access.

How ChatGPT decides which brands to recommend

We tracked 12,000 buyer prompts across four engines. Three signals explain most of who gets named.

Maya OkaforHead of SearchSep 18, 2026 · 7 min read
Key takeawaysRewrite your top five pages so the answer comes firstPublish one comparison page against your closest rivalAdd Organization schema and an llms.txt fileMake your pricing quotable in one sentence

Engines shortlist. They don't rank.

Google gives you ten links. ChatGPT gives you three names. Position four used to mean some traffic. Now it means none.

Across 12,000 prompts, the top three brands in an answer took 91% of follow-up clicks. Being fourth is the same as not being there.

Signal 1: a clear, quotable answer

Engines lift sentences, not pages. Pages that answer the question in the first 40 words were cited 3.4 times more often than pages that got there eventually.

If an engine can't quote one sentence from your page, it will quote someone else's.

Signal 2: comparison content

When buyers ask for the best tool, engines lean on pages that compare options. Brands with a comparison page against their main rival appeared in 62% more shortlists.

Signal 3: the same facts everywhere

Engines cross-check. When your pricing, category and positioning match across your site, docs, reviews and directories, you get named more often. When they conflict, engines hedge by naming someone else.

What to do this week

  • Rewrite your top five pages so the answer comes first
  • Publish one comparison page against your closest rival
  • Add Organization schema and an llms.txt file
  • Make your pricing quotable in one sentence
Written by Maya OkaforHead of Search at Scaplet

Keep reading

Put this playbook to work.

Free audit. No card. No follow-up call, promise.