How to report AI search visibility to a board
Report four things: share of voice against a fixed competitor set, brand accuracy, source concentration, and citation rate per platform. Never a blended score, modelled traffic shown as measured, or a projection.
The four numbers a board will accept
Share of voice, against a competitor set fixed in advance
Boards understand relative position. This is the closest analogue to market share they will find in this data, and holding the competitor set constant is what makes the series credible.
Brand accuracy
The share of answers naming you that contain no material error. This is the one number in the set that maps directly onto risk, and it is the one non-marketing directors engage with fastest.
Source concentration
Which domains carry citations in your category. It explains why budget is going where it is going, and it pre-empts the question about why you are spending money on Reddit.
Citation rate, per platform
Four or five numbers rather than one. Boards tolerate this better than marketers expect, provided you say which platform matters most and why.
The three that will get you challenged
A blended AI visibility score
It is the most saleable number in this category and the first thing a numerate director will ask about. Once you explain it averages platforms that behave differently, its credibility is gone and it takes the rest of the pack with it.
AI-attributed traffic or pipeline
Most assistant surfaces send no referrer. Anything beyond genuinely referred clicks is modelled, and presenting a model as a measurement is the fastest way to lose a reporting line permanently.
Projected visibility
Forecasting citation rates on platforms that change without notice is guessing with a chart attached. Boards have seen a lot of charts.
How to handle the attribution question
It will be asked, usually by the CFO, usually as "what did this produce?" The answer that survives is the honest one, delivered before the question rather than after it.
Split the report into three explicit buckets. Measured: Perplexity referrals and any other genuinely referred clicks, which are small and real. Modelled: your estimate of assistant-influenced pipeline, with the model shown. Unknown: the share of direct and unattributed traffic you believe is assistant-influenced but cannot demonstrate.
Then say plainly that the third bucket is growing, that this is an industry-wide measurement problem rather than a failure of your instrumentation, and that the metrics above exist precisely because attribution on these surfaces is not available.
Our position
A board will forgive an honest unknown. It will not forgive discovering that a confident number was modelled, and the discovery usually happens in the meeting rather than before it.
A one-page structure
- 1One line on what changed this period, in plain language, before any numbers.
- 2Share of voice against the fixed competitor set, as a trend rather than a point.
- 3Brand accuracy, with the specific errors listed and their source named.
- 4Citation rate per platform, with one sentence on which platform matters most for your buyers.
- 5Source concentration, as the justification for where the work is going.
- 6The measured, modelled and unknown split for anything traffic-related.
- 7What you are doing next, and what would make you stop.
What to say when the numbers are bad
Low visibility in a new measurement discipline is not a performance failure, and framing it as one invites the wrong conversation. It is a baseline.
The useful framing is diagnostic: here is where we are absent, here is which of four causes is producing it, here is what each remedy costs and how long it takes. A board responds well to a diagnosis and badly to a number with no explanation attached.
If the cause is off-site absence, say so early and set expectations on timescale. Community and publisher work runs over quarters, and a board expecting quarterly movement from it will conclude the programme failed at month four.
Related questions
The definitions behind every number we report
Related answers
What is AI visibility?
AI visibility is how often, and how accurately, a brand appears in AI-generated answers.
Read the answerHow-toHow to monitor AI search visibility
Build a prompt set of 30 to 50 buyer questions and freeze it.
Read the answerDecisionDoes GEO need its own budget, separate from SEO?
Partly.
Read the answer