The problem
The client asks why they never come up in ChatGPT, and you have no instrument to answer with. Without a baseline, nothing you do about it can be shown to have moved anything.
When to use this playbook
A client asks about AI search, or you start content or technical work aimed at it.
The workflow
- 01
Start from the Diagnostic
The Discoverability pillar reads whether the site’s robots.txt lets AI retrieval bots in, whether the server turns away crawlers the rules invite, and whether structured data is published.
- 02
Run a Citation Scan
Ten fixed questions about the client’s category, asked across six AI answer engines: sixty answers. You see every domain each answer cited and a ranking of how often each appears, with the client and any competitors you name in it.
- ChatGPT
- Claude
- Gemini
- Perplexity
- Google AI Overview
- Google AI Mode
- 03
Check that AI crawlers can reach the pages
If the Diagnostic shows crawlers being turned away, a Crawlability Scan reads the client’s pages URL by URL and shows which ones AI agents and crawlers are refused.
- 04
Make the changes
Content, structured data, robots.txt rules, server settings: the work is yours. The findings say where to start.
- 05
Re-scan and compare
A re-scan of the same domain, category and country asks the same ten questions, so the two scans compare. Citation Scan Trends shows the movement over time. Or put the pair on the Citation Scan Watch: the same scan every month for $19 USD, cancel anytime.
What you use
- Diagnostic Ten pillars scored from public evidence in 5–10 minutes, with every finding turned into an action item.
- Citation Scan Sixty AI answers to ten fixed questions about the category, and every domain they cite.
- Crawlability Scan URL by URL, which pages AI agents and crawlers can reach and which they are turned away from.
What you end up with
A before and after of how often AI answers cite the client, against the same competitors, on the same questions.
The opportunity
AI-visibility work with a measurement attached, rather than a promise.
What these scans do not do
- A Citation Scan observes sixty answers on one day. AI answers vary from one run to the next, so read small movements with care.
- A Crawlability Scan is a reading at one point in time. It does not compare one scan with the next.
What comes next
Prepare for a client review
Pull what was wrong, what was done, what changed and what is still open for one client over any period, from the work already in their workspace.
Read the playbook →Find new work in existing accounts
Put every client on a scheduled Diagnostic and let what changes between runs show you where the next piece of work is.
Read the playbook →