4 min read
How to report AI visibility to clients
A step-by-step reporting outline for agencies: mention rate, citation rate, competitor gaps, page outcomes and AI-referred traffic, using RootSmith's checks.
Report AI visibility to clients monthly. Show each rate beside its run count, split by engine and compared with the prior month, then add competitor gaps and page outcomes. RootSmith supplies the checks and the 14, 30, 60, and 90 day page measurements. Label traffic figures as a lower bound, and never promise citations.
Before you start: what you need
Three things need to be in place before the first report.
- Client authorization. RootSmith is for business and professional use. Agencies may manage client sites with the client's authorization, so get that in writing first.
- An agreed scope. Settle the buyer questions to track, the AI assistants to cover, and the reporting cadence.
- Analytics access. You'll need to see AI-referred sessions and Google clicks.
Set the cadence at monthly. Run-to-run variance in AI answers makes daily movements mostly noise.
Step 1: Pick the tracked buyer questions and engines
Start with the questions the client's buyers ask AI assistants when choosing a vendor or tool.
RootSmith tracks what ChatGPT, Gemini, Perplexity, Claude, and Google AI Overviews say for those questions, including competitor mentions and citation rates. Every tracked question gets a full check each week and a lighter check on the other days.
For Perplexity and Claude, RootSmith asks their APIs with web search and labels those results. Put that in the report's method note so the client knows how each engine was checked.
Step 2: Report mention rate and citation rate with their run counts
Report both rates for each engine. A mention means the client is named. A citation means a page is linked as a source. The AI mention vs AI citation article covers the difference.
Put the number of runs next to every rate. "35.3% on 2,940 runs" is a claim, while "35.3%" alone is decoration.
Show a dash instead of a zero for an engine with too few runs. A zero says the client never appears, and thin data can't support that.
Leave out raw mention counts without denominators, blended cross-engine scores, and anything measured once.
Step 3: Compare month over month and state what counts as change
Compare each month with the one before. If a change sits inside the margin of error, write it as unchanged.
Tell clients plainly that no one can honestly guarantee citations, because AI answers change from day to day. RootSmith works on the same terms: it shows where you stand, helps you publish better pages, and shows what changed afterward.
Step 4: Show competitor gaps
List the tracked questions where a competitor is named and the client is not. Flag a competitor moving ahead in the answers as an item that needs action.
Pair each gap with the page you plan to write or refresh. The guide on stopping ChatGPT recommending competitors is a useful companion.
Step 5: Report page outcomes
RootSmith checks each published page's AI citations and Google clicks at 14, 30, 60, and 90 days. Use those four checkpoints as the rows of the page outcomes section, so the client sees which pages worked.
If a page the engines used to cite no longer appears, raise it with the client.
Pages publish on the client's own domain, so the results build on the client's site.
Step 6: Add AI-referred traffic, labelled as a lower bound
Count a session as AI-referred when its source is a known assistant host or tag, or when GA4 puts it in the AI Assistant channel. A session that meets both counts once.
Links ChatGPT cites often carry utm_source=openai, and Gemini has been seen adding utm_source=gemini.
Clicks from Google's AI Overviews and AI Mode arrive as ordinary Google organic. Search Console's Generative AI report shows impressions there, not visits.
Call the final number a lower bound, because analytics attributes some AI referrals to direct. Use those exact words in the report.
Step 7: Close with a one-page summary and next step
Answer three questions:
- Is it working? Give one visibility figure with its range, run count, and the month it's compared with.
- Is it worth the money? Give AI-referred conversions or revenue, labelled as a lower bound.
- What happens next? Write one line.
Raise a flag when visibility falls against its own recent past, a competitor moves ahead, a page the engines cited drops out, or conversions fall in the same weeks as visibility.
See what AI says about you
Start free. Pick a plan when you're ready for tracking and content.