What a client-ready summary contains
| Section | What it shows | Why the client cares |
|---|---|---|
| Mention rate | Share of answers that name the client, per engine | The headline: are we in the answers at all? |
| Competitors | Who is named instead, and how often | Clients understand “they get named, we don’t” instantly |
| Cited pages | The pages the answers relied on, each with an action | Turns the problem into a to-do list |
| Per-question view | Each question and what each engine answered | Lets the client read the evidence themselves |
| Method | Engines, number of questions and runs, date | Makes the numbers credible and repeatable |
See all of these in a real format in the example report, a complete full report for a brand that didn’t sponsor it.
Choosing the questions with the client
- Ask the client how their customers describe the problem, in their own words.
- Include the place, budget or situation where it matters (“for a small clinic”, “in Manchester”).
- Keep brand names out of the questions; you want to know whether the client is named unprompted.
- Agree the list before running it, so nobody can say the questions were picked to look good or bad.
Presenting it without overpromising
- Say that answers vary between runs, and show how many runs the numbers are based on.
- Don’t promise a position. Nobody outside the AI companies controls who gets named; promise the work on the cited pages and the client’s own site, and measure the result.
- Link each recommended action to a cited page, so the plan is grounded in what the answers actually used.
Run it
Start with the free AI visibility check for a prospect. For a client, the full report runs 15 questions 2 times on OpenAI and Claude with web search, plus Gemini from model knowledge, for a one-time $9, with one re-check after 30 days. For reading the cited pages with the client, see which pages ChatGPT cites.
Showing clients their AI visibility: FAQ
Can I send a client a screenshot of a ChatGPT answer?
You can, but one answer proves little: run the same question again and the list can change. Show how often the client was named across several runs and engines, and keep the screenshot as an illustration, not the result.
How do I share a CheckAIVisibility report with a client?
Every report has a link that anyone with it can open, for 90 days. Send the link, or walk the client through it on a call.
How do I show progress after we’ve done the work?
Run the same questions again. The full report includes one re-check after 30 days on the same questions, so the client sees a before and after on a like-for-like basis.
What should I tell the client the report doesn’t cover?
That it uses the official APIs of OpenAI and Claude with web search and Gemini from model knowledge, not the consumer apps, and that it doesn’t test Perplexity or Google AI Overviews. Saying this up front avoids a client finding a different answer in the app and doubting the whole report.
Is there a free way to start the conversation?
Yes. The free check runs 5 questions on OpenAI with web search and shows the same sections in smaller form. It is often enough to show a prospect that a gap exists.