Home / Question map
Free tool updated 2026-11-22

Which buyer questions does your site answer?

Enter your domain. We propose the questions people are likely to ask AI about you, from your own pages, and check each one the way an AI crawler reads your site.

How it works

  1. Propose. We read your homepage, robots.txt and sitemap and list up to 10 questions your pages suggest people ask.
  2. Check. Each question goes through the question doctor: is there a page for it, does it answer plainly, can AI crawlers read it?
  3. Fix the worst first. Missing pages and crawl problems come first: there's nothing there a crawler can quote. A weak page may need only a heading that asks the question and one sentence that answers it. Open any question for its full diagnosis.

Same check: earncitations map <domain>, the MCP tool question_map and GET /api/map (docs).

Questions

Where do the questions come from?

From your own site. For software: what it costs, what it is, a free trial, then what your pages show you care about: an API if you publish developer docs, SOC 2 if you have a security page, a "vs" question for each comparison page you have, the integrations you list (well-known tools first), and a "best X for Y" question if your title or main heading claims one. For hotels: check-in, check-out, parking, pets, breakfast, cancellation and room price. For restaurants: hours, reservations, vegetarian options, takeout, parking and location.

Why start from questions instead of keywords?

Because people ask assistants whole questions before they choose, and a page that answers the question plainly is the easiest thing for an assistant to quote. The loop Granola published (summarised by Jam, spreadjam.com) starts there: list the buyer decisions, check which ones you lose, diagnose why, fix one thing, rerun. This page does the first two steps for free.

What do the verdicts mean?

Answered: a page of yours answers it plainly; what's left is off-site. Weak: a page covers it but doesn't answer plainly (no heading that asks it, no direct first sentence, no price or time). Missing: nothing on the site covers it. Crawl problem: the page exists but AI crawlers can't read it. If your homepage refused our crawler (possibly bot protection), we say we couldn't read the site rather than guess what AI crawlers get. Open any question for the full diagnosis.

Does this ask ChatGPT?

No. It reads your pages the way an AI crawler does and makes no AI calls, so it's free and repeatable. Whether assistants already cite you is a separate question: ask them, and use the off-site source list in each diagnosis.

Can I follow these questions over time?

Yes. Download the map as a question set and load it in the answer log. Ask each question on the assistants you care about, check each answer with the Answer check, and save it. Each question keeps its type as a topic (pricing, integrations, comparisons…), so the log shows how often you're named per buying decision, and which competitor is named ahead of you in each.

Can I map my own questions?

Yes: earncitations map yoursite.com -q "…" -q "…", GET /api/map?url=…&q=…&q=…, or the MCP tool question_map with questions. To be told when an answer breaks later, add them to watch with -q.