“You are missing from AI answers” is an effective way to start a sales conversation. It is not, by itself, a reason to commission a large audit.
The useful question is what happens if you appear. Does the answer reach someone with a relevant problem? Is there a page that helps them evaluate you? Can they take a sensible next step? Can your team tell whether the resulting enquiry is worth pursuing?
Answer engine optimisation, or AEO, can describe legitimate work on discoverability and useful content. The label should not allow a supplier to skip the commercial reasoning that any search project needs.
Begin with a buying question, not a prompt collection
Suppose you sell a specialist implementation service. A broad question such as “What is the best agency?” could reach almost anyone. A question about migrating a complex content model while preserving editorial workflows describes a much more specific situation.
Build your research around the decisions customers actually make: recognising the problem, comparing approaches, checking suitability and reducing the risk of choosing a supplier.
Talk to the people who handle enquiries. Review the questions that arrive before a proposal and the objections that delay a decision. Those are better starting points than producing hundreds of near-identical AI prompts because a tool makes it easy.
Trace one complete route
Here is a hypothetical route for a migration enquiry. It is a planning example, not a claim about measured results.
- A buyer asks how to leave a headless CMS without making the marketing team dependent on developers.
- A useful article explains the trade-offs and links to a relevant migration service.
- The service page shows the scope, examples and the decisions that affect the work.
- The buyer can book a discussion or send a brief without navigating an unrelated offer.
- The enquiry reaches the right person with enough context for a useful response.
The missing piece might be search visibility. It might also be an unconvincing service page, an inaccessible form or a slow follow-up. An audit that examines only the first step cannot distinguish those problems.
Separate visibility, visits and business outcomes
These measures answer different questions. Visibility indicates that material appeared. Visits show that someone arrived. An accepted opportunity reflects a business judgement about fit and potential.
Google now documents a Generative AI performance report in Search Console covering impressions in AI Overviews and AI Mode. Availability can depend on sufficient activity. This is a visibility measure, not a pipeline report and not a view of someone’s complete buying journey.
For on-site measurement, Google Analytics distinguishes lead-generation and lead-qualification events. Your organisation still needs to define what qualification means and connect reporting appropriately.
Keep the limitations visible. A prospect may read an answer, return directly days later and mention a colleague’s recommendation. No single attribution field explains all of that. Report the evidence you have without claiming precision it cannot support.
Ask for a page-level improvement plan
A useful recommendation identifies the page, the buyer’s unanswered question, the evidence needed and the next step the page should support. “Add more FAQs” is not enough.
For a service page, that could mean explaining what a migration includes, showing an actual project and making the enquiry action easier to find. For an article, it could mean adding an original comparison, correcting a stale claim or linking to a genuinely relevant service.
Google’s guidance on AI features points back to crawlability, index eligibility and useful accessible content. Special markup should not become an excuse to avoid the work of making the page convincing.
What a credible engagement should deliver
- A prioritised set of buyer questions tied to real services and commercial fit.
- A baseline with dates, definitions and the limitations of each data source.
- Specific page changes with an owner and a reason for their priority.
- A measurement plan that follows meaningful actions beyond impressions.
- A review point to decide what to continue, change or stop.
Ask whether the supplier will help implement the findings. A large export can be useful research material, but it is not the same as improving the experience your next customer will have.
Make the first investment small enough to learn from
Choose one service with a clear offer, a relevant audience and a working enquiry route. Improve a small group of connected pages, then review visibility, engagement and lead quality over an appropriate period.
Do not promise a fixed number of AI citations or a guaranteed pipeline uplift. Search behaviour changes, answers vary and sales depend on more than the website.
If that is the kind of work you need, our SEO and conversion service connects discovery with the on-site journey. If implementation is the ongoing bottleneck, read what a fractional marketing engineer does. The deliverable should be a better route to a customer, not just another score.
