How to get your product recommended by ChatGPT: a playbook

More buyers now ask an AI assistant which tool to pick before they ever search Google. If ChatGPT names three competitors and not you, you've lost that buyer without knowing they existed. This playbook is the process we recommend to founders who want to change that. It's practical and specific, and it doesn't involve gaming anything.

Step 1: write down the questions your buyers actually ask

Forget keywords for a moment. Write the questions a buyer would type into ChatGPT when they're about to choose a product like yours. Good questions have buying intent and some detail:

• "Best invoicing tool for freelancers in the UK"
• "Harvest alternatives that are cheaper"
• "Which time tracker works with Linear?"
• "Is [your brand] good for small agencies?"

Aim for 10 to 20 questions across three kinds: category questions ("best X for Y"), alternative questions ("X alternatives", "X vs Y") and problem questions ("how do I do Z"). Include the long, specific ones your best customers ask. As our 20-brand test showed, smaller brands rarely win the broad questions but can win the specific ones.

Step 2: measure where you stand

Put each question to each engine and read the full answers. Note whether you're named, in what position, who's named instead, and which pages are cited. Doing this by hand is fine for a one-off. To see changes over time you need the same questions asked on a schedule; that's what Citemark does, and the free check gives you a first snapshot in a minute.

Step 3: read who wins, and why

For every question where you're missing, read the answer and ask:

1. Who was named? Are they real competitors, bigger brands, or adjacent tools?
2. What reasons did the answer give? Price, an integration, a free plan, a specific feature? Those reasons are how the engine sees the category.
3. Which pages were cited? These are the sources the answer was built from.

Often the pattern is obvious once you read five answers in a row. The winners have a comparison page that the engine quotes, appear in a popular "best tools" list, or come up in a Reddit thread where people recommend them.

Step 4: get onto the pages that get cited

The cited pages are your most concrete lead. They fall into a few types, and each has its own approach. Which websites AI engines cite covers them in depth; in short:

• "Best of" and comparison articles: contact the author with a short, honest note on what you do better and for whom. Offer facts they can check. Many lists are updated yearly.
• Review sites: ask happy customers for reviews. A product with no reviews rarely makes a shortlist.
• Community threads: take part where your buyers already ask for recommendations. Be useful, say who you are, and don't spam.
• Your competitors' pages: if their comparison page is cited, publish a fair one of your own.

Step 5: publish the page the answer is missing

Sometimes the engine cites nothing useful for a question, or cites a page that's out of date. That's a gap you can fill. Useful pages to have:

• A comparison page for each main competitor ("[You] vs [Rival]") with an honest table of differences.
• An alternatives page ("[Rival] alternatives") that explains who should pick what.
• Use-case pages for the specific buyers you serve best ("for agencies", "for freelancers").
• A pricing page with real numbers. Answers often mention price; give them the right one.

Write these for humans first. Pages that are clear, specific and fair are the ones that get quoted.

Step 6: make your own site easy to quote

AI engines summarise. Make sure the summary of your site is right:

• Say what you do and who it's for in the first lines of your homepage, in the words your buyers use.
• Put your price in plain text, not only in an image or a calculator.
• Keep one consistent description of your product across your site, directories and social profiles.
• Use headings that match questions people ask, and answer them directly underneath.

Step 7: check that AI crawlers can read your site

If your robots.txt blocks the crawlers that AI search uses, your pages can't be fetched for answers. The ones to check first are OAI-SearchBot (ChatGPT search), ClaudeBot (Anthropic) and Google-Extended (Gemini). The free AI SEO audit checks your robots.txt and homepage in seconds. Some sites also publish an llms.txt file, a proposed plain-text summary for AI tools; the llms.txt generator can draft one. It's cheap to add, but don't expect it to change answers on its own.

Step 8: track, then repeat

Before you start changing things, write down your baseline: your score, the questions where you're missing, and the pages cited for each. Without a baseline you won't know which change made the difference.

Changes take time to show up: new reviews have to be indexed, articles updated, models refreshed. Track the same questions weekly and look at the trend after 4 to 8 weeks. When a question moves, look at what changed in the cited sources. When nothing moves, read the answers again and try the next lever.

Citemark's dashboard is built around this loop: daily answers, the rivals named instead, the sources cited, and a Fixes list per question. You can start with a free check and see your first gaps in about a minute.