The question clients are asking
"Does ChatGPT recommend us?" is becoming a normal question in client meetings. It is a fair one.
When a customer asks an assistant for a recommendation, they get one written answer that names a few businesses, sometimes only one. If the client is not in it, they may never be considered.
Most agencies have no settled way to answer yet. Screenshots of a single answer do not hold up: the client asks again, gets a different answer, and the report looks wrong. What works:
- A fixed set of questions, asked repeatedly.
- The result written as "named in X of Y asks".
- A list of the work done.
What you can do with GEO Agency
- Score the pages that matter. Page analysis scores a page across seven areas (answerability, verifiability, structure, freshness, authority, crawlability and safety) with the reasons and suggested fixes.
- Find the weakest pages first. Site audits crawl up to 100 public pages and rank the weakest, so the work starts in the right place.
- See how an answer might look. Answer simulation shows how an assistant might answer a question and whether the client's page is likely to be used.
- Spot gaps against rivals. Compare a client page with a competitor's and list the questions it does not answer yet.
- Report under your own name. White-label PDF reports and share links on paid plans.
- Work as a team. The Agency plan adds a shared team workspace and a WordPress connection that pushes approved changes. Nothing is published until a person approves it.
A simple monthly routine
- Agree five to ten questions with the client, in their customers' words.
- Take a baseline: check how assistants answer today and score the pages that should answer.
- Fix the likely causes, most important first: missing answers, weak structure, missing schema markup, thin or out-of-date content.
- Re-check the same questions the same way each month and send one plain report.
Our guide to monthly AI visibility reporting for clients has a report template, wording to avoid and a printable checklist.
Plans
- Free: 3 credits a month, no card.
- Pro: £29 a month (VAT included), 50 credits, site audits, answer simulation, PDF export, share links and white-label reports.
- Agency: £129 a month (VAT included), 500 credits, plus the team workspace, WordPress connection and the AI agent workspace.
Prices are VAT included and you can cancel from Billing at any time. A page analysis costs 1 credit and a site audit 10. Full details are on Pricing.
Honest limits
- The tools run on Google Gemini, so scores and simulations are estimates, not a record of what ChatGPT, Perplexity or any other assistant said to a customer.
- No promises. Nobody can promise a client a place in an AI answer, and we do not.
If you would rather we ran the checks and fixes for a client, the done-for-you service is quoted per piece of work.
Common questions
Can I put my agency's name on the reports?
Yes. White-label PDF reports are included on the Pro and Agency plans. If you want us to run the checks for a client under your name, ask about the done-for-you service.
Which plan suits an agency?
Agency, at £129 a month with VAT included, if several people work on several clients: it adds a shared team workspace, the WordPress connection and 500 credits a month. Pro, at £29 a month (VAT included), suits a consultant or one client at a time.
Does it track ChatGPT for my clients?
No. Analysis and answer simulation run on Google Gemini, so results are estimates of how an assistant might answer, not a log of what ChatGPT said. Say so in your reports; clients respect it.
Can we try it before paying?
Yes. The free plan has 3 credits a month and needs no card, enough to analyse a few pages and see the reports.
Can we promise clients a place in AI answers?
No, and we would advise against it. Nobody controls what assistants say. Sell the checking, the fixes and an honest measure of change.