AI Search Visibility

Be understood
in AI search.

We review how your business appears in a selected set of AI search answers, find the gaps in what’s known about you, and improve the content, evidence and website foundations you control. You get a documented baseline, a prioritised plan and the agreed work done.
Book a discovery call
See what we assess

What do the answers say about your business?

When someone asks an AI tool who can help with a problem like the ones you solve, three different things can be true.

You’re mentioned.

Your name appears in the answer. That’s all it means.

You’re described correctly.

The services, locations and capabilities in the answer match the business you run today.

The answer can be checked.

It points to sources, yours or someone else’s, that support what it says.
A business can be mentioned and misdescribed. It can have real expertise that’s hard to pick up from its public content. Its website and an old directory listing can disagree. And sometimes a customer says they asked an AI first, and you have no idea what it told them.
One missing mention isn’t proof of lost business. It’s a reason to look properly.

The questions, the answers and the sources behind them.

We follow the trail backwards, from what a buyer asks to what your own content says.
  • The questions
  • The answers
  • The sources
  • Your content and access
01

The questions

The research and comparison questions that matter for your priority services, audiences and locations. Some name you and some don’t, and we keep the two apart. Being described well when someone asks for you by name says nothing about being discovered when they don’t.
02

The answers

What selected AI search tools say in response. That might include Google’s AI features, ChatGPT search or Perplexity. We agree the set for your engagement and record each answer with its question, platform and date.
03

The sources

What those answers cite or link to, on your site and elsewhere, and what each source actually supports.
04

Your content and access

Whether your public information is clear, consistent and backed by evidence, and whether the relevant systems can reach it under the settings you choose. Those settings can be precise. OpenAI, for one, runs separate crawlers for search and for model training, so being eligible for ChatGPT search doesn’t mean offering your content for training. Eligible isn’t included, though, and no file, tag or page switches inclusion on.

Findings you can act on.

Four things, in order: a dated baseline of what we observed, prioritised findings, the agreed improvements carried out, and a follow-up check against the same questions.

Every finding says what matters, why it matters, who can change it and what happens next. The work that follows might be a corrected service description, stronger project evidence, a clearer answer to a common buyer question or a fix to an access problem. We recommend a change because it closes a specific gap. Not every page needs an FAQ, and not every business needs more articles.

We can change what you control. We can ask an outside publisher to correct a listing. We can’t make a publisher or an AI provider change anything, and we won’t pretend otherwise.

One entry from a findings register

Made up to show the format. No result is claimed.
Asked
“Who provides [service] in [region]?”
The answer said
the business serves an area it left two years ago.
What we found
the business’s own location page and an old directory listing disagree.
Who can fix it
the business, for its own page. The directory, on request.
Next
correct the location page, request the directory update, then re-check the same question.

Connected to your SEO and content work.

If you already invest in SEO, most of the foundations should be in that plan: pages that can be crawled, clear service content, accurate business details, structured data where it applies. For Google’s AI features in particular, Google’s own guidance is that ordinary good practice is what counts.

What this service adds is the looking.

A defined set of buyer questions put to selected AI tools, a dated record of what comes back, a check of the sources behind it and a plan for the gaps. Any improvements join one prioritised list with your SEO and content work, so you never pay for the same fix under two names.

Sometimes the finding is that your existing plan needs adjusting, not that you need another retainer. If so, that’s what we’ll recommend.

SEO

What an observation can tell you, and what it can’t.

We track a defined sample of questions, the work completed and identifiable website activity where it exists. Then we’re careful about what each observation means.

A mention

Shows
you appeared in that answer.
Doesn’t show
your reach across all users or all questions.

A citation of your site

Shows
the answer linked to your page.
Doesn’t show
that it endorsed you, or relied on that page alone.

A recommendation

Shows
you were offered as an option in that context.
Doesn’t show
a stable ranking, or the same answer next time.

A referral you can track

Shows
a visit or enquiry happened.
Doesn’t show
the full influence of AI on that buyer. And no referrals doesn’t mean nobody saw you.
We keep a stable core set of questions and report exploratory ones separately, so a changed sample can’t pass as an improvement. Any proportion comes with the number of checks behind it, as a result within the sample, never as market share. A before-and-after difference is something to investigate, not proof that one edit caused it.

The honest position

This is a young field, and we’ll be straight about that. We don’t yet have AI search case studies, and we won’t relabel SEO results to fill the gap. What we can show you is the method, a sample of the deliverable, and 20 years of the content, search and website work it depends on.

Start with a defined review.

The sensible first engagement is a bounded review and a prioritised set of improvements, with the platforms, questions and pages named in the scope. It stands on its own.

Ongoing monitoring has to earn its place. It’s worth paying for when it will change the work or inform a decision, at intervals that suit how much is changing. If it won’t, we’ll tell you not to buy it.

The work can also sit inside a Growth Partnership, alongside search, content and website priorities. And if the bigger problem is unclear, a Clarity Audit is available, not required.

Practical questions

Is this different from SEO?

Yes, in what it investigates. SEO works on being found and on the pages people land on. This adds a review of AI answers and the sources behind them. Shared foundations are planned once.

No. Nobody outside those platforms controls their answers. We can investigate how you appear and improve the information you control.

Answer engine optimisation and generative engine optimisation. They’re industry names for roughly this kind of work. The scope matters more than the acronym.

The ones your buyers are likely to use, agreed and written into the scope. We don’t claim to watch every assistant in every market.

By comparing a fixed sample of questions over time, alongside accuracy findings, completed work and trackable visits or enquiries, with the limits stated.

No. This is about your public search presence, not chatbots, workflow automation or model training.

Yes. Implementation and any ongoing review are agreed after the findings.

Ask an AI about your business.
Then talk to us.

One answer proves nothing, but it’s a good place to start. Bring what it said, and how your customers usually research a decision like yours. 

Book a discovery call