Your business appears in an AI answer. That is useful to know, but did it produce an enquiry, a qualified opportunity or a sale?
Measure GEO with three separate groups of evidence: visibility in answers, identifiable website visits and commercial outcomes. Connect them where the data allows, but do not present a brand mention as a lead or assume every unexplained enquiry came from AI.
This guide is for business owners who want a practical reporting method without a dashboard full of impressive but disconnected numbers.
Start with the business outcome
Define what a worthwhile result looks like before collecting more data. For a consultancy, that could be a qualified discovery call. For an installer, it could be a survey booking within the service area. For a manufacturer, it might be a specification request from a suitable buyer.
Record the difference between a raw enquiry and a qualified one. Spam, job applications and requests for a service you do not provide should not count as evidence that your GEO campaign is working.
Use a baseline period that reflects your normal sales cycle. Seasonal demand, advertising changes and a new referral partnership can affect enquiries alongside your website work.
1. Track visibility without calling it revenue
Create a consistent set of commercially relevant questions covering discovery, comparison and purchase decisions. Include the sectors, locations and customer requirements you actually serve.
For each observation, record:
- The exact question and date.
- The AI platform and whether search was used, where visible.
- Whether your business was mentioned.
- Whether your website was linked or cited.
- Whether the description was accurate and relevant.
Repeat observations over time and keep the question set documented. A larger number of mentions can reflect a changed test set rather than genuine improvement.
If you report a mention rate, show its denominator: “Mentioned in 8 of 40 recorded answers” is more informative than an unexplained “20% AI visibility score”. Those observations are not necessarily independent, so avoid attaching false statistical certainty.
2. Inspect identifiable AI referral visits
In your analytics tool, examine referral sources from AI services and group known sources where appropriate. Check actual source values rather than assuming every product sends the same referrer.
Validate the measurement with a small number of controlled visits where practical, and distinguish those checks from customer activity.
Not every AI-assisted journey produces an identifiable referral. Someone may copy your business name, search for you later or switch devices. Apps and privacy settings may omit referral information.
Likewise, a referral from an AI service is evidence of a visit, not proof that a particular page edit caused it. Keep analytics limitations visible in the report.
3. Connect visits to enquiries carefully
Configure meaningful events such as a successfully submitted enquiry or a confirmed booking. Counting every button click as a lead will inflate the result, especially if visitors click more than once or a form fails.
Test that events fire once and only at the intended step. If the booking takes place on another service, document whether your reporting can follow that journey.
Where your consent arrangements and applicable rules permit, record relevant source information with the enquiry. Do not place names, email addresses or message contents into analytics events or tracking URLs.
For a wider discussion of finding and qualifying demand, see our AI lead generation guide.
4. Ask customers how they found you
A short optional question can reveal journeys that analytics misses: “How did you first hear about us?” Allow a free-text answer alongside simple choices.
If someone selects an AI assistant, a follow-up on a sales call can clarify which one and what they were researching. Do not coach them into crediting AI.
Keep self-reported sources separate from measured referrals. A buyer may have seen a recommendation, read a directory and then searched Google; there is no need to force that journey into a single supposedly perfect attribution label.
5. Join the lead to its commercial outcome
Use your CRM or a simple controlled spreadsheet to record enquiry date, source evidence, qualification, opportunity stage and eventual value.
For example, an AI referral might lead to one enquiry that becomes a qualified opportunity but has not closed. Report exactly that. Do not count the full potential contract value as revenue.
An illustrative monthly summary might contain:
| Measure | Example result | What it tells you |
|---|---|---|
| Recorded answers mentioning the business | 8 out of 40 | Visibility within that test set |
| Identifiable AI referral visits | 24 | Measurable referral traffic |
| Enquiries linked to those visits | 2 | Observed website response |
| Qualified opportunities | 1 | Commercial relevance |
| Confirmed sales | 0 so far | No closed revenue yet |
These numbers are an example, not a benchmark or a Panovista client result.
6. Review changes without claiming causation too quickly
Keep a dated record of content updates, technical fixes and external coverage. Review visibility, visits and leads together after enough time for your sales cycle.
Small samples fluctuate. A month with two enquiries followed by one with four is not enough on its own to prove that a content change doubled performance.
Our guide to why ChatGPT recommends competitors shows how to investigate visibility gaps before deciding what to change.
Questions business owners ask
Can an AI mention help without a website click?
It may influence awareness or later research, but the commercial effect is harder to establish. Record what you can observe rather than assigning speculative revenue.
Which number matters most?
Qualified opportunities and confirmed sales matter commercially. Visibility and visits are useful diagnostic measures, particularly while the sales pipeline is still developing.
What should I ask a GEO agency to report?
Ask for the prompt set, dated observations, measurement limitations, identifiable referrals and qualified outcomes. Contact Panovista to discuss a reporting plan matched to your sales process.