AI visibility audit

AI visibility audit: what ChatGPT tells your next customer about you

A fixed-fee, manual audit of what ChatGPT, Google AI, Perplexity and Claude say when a customer asks about a business like yours, scored across the four engines with a ranked fix plan. For any UK business whose enquiries start online, run from London, £500 at the launch rate and £750 standard, report in five working days. The next step is a 20-minute call.

Book a 20-minute call What is checked →

Fixed fee · launch rate

£500~~£750~~

Fixed fee, report in five working days. Three audits a month at this rate.

4AI engines: ChatGPT with search, Google AI, Perplexity and Claude

10Real customer prompts, fixed wording, run under controlled conditions

5Scored dimensions, blended into one headline score per engine

5Working days from go-ahead to the report in your inbox

The question

What is AI search visibility?

AI search visibility is whether your business appears, and how it is described, when someone asks an AI assistant a question you would once have answered from a Google results page. Ask ChatGPT for a chartered surveyor in Watford, or Perplexity for a physiotherapist in Bristol who treats runners, and you get one answer with three or four names in it, not ten blue links. Being in that answer, described accurately and without a caveat, is visibility. Being absent is not a ranking problem you can see in your analytics, because the enquiry never reached you.

It rests on different evidence from search rankings. The engines draw on what your site says in plain text, whether the same facts about you appear consistently on directories, reviews and LinkedIn, whether your pages answer the questions people actually ask, and whether anyone else has written about you. A business can rank well on Google and still be invisible, misdescribed or recommended against inside AI answers. The audit measures that directly instead of guessing from rankings.

The method

How does the audit work?

An AI gives one answer, not a list. If that answer names a competitor, states your prices wrongly or attaches a caveat to your name, you lose enquiries without ever seeing them. Your analytics cannot show you this. A person running real questions and reading every answer can.

1
Ten real customer prompts

Discovery ("best chartered surveyors in your town", "a marketing agency in Leeds for a B2B firm"), reputation ("is this business any good?"), commercial detail ("what do they charge?") and head-to-head against your main rival. The wording stays fixed, so audits are comparable.
2
Four AI engines

ChatGPT with search, Google AI, Perplexity and Claude, each run under controlled conditions with UK location and a fresh session per prompt. Every answer screenshotted as evidence.
3
Five scored dimensions

Presence, competitive position and source basis are scored here; accuracy and sentiment are scored under brand visibility below. Each is 0 to 4 per engine against a written rubric, blended into a single headline score.

By name

AI brand visibility: what the engines say when someone asks about you by name

AI brand visibility is the second half of the audit. Discovery prompts test whether you are found; brand prompts test what is said once you are. A prospect who has your name from a referral will ask an assistant whether you are any good, what you charge and who else they should consider, and the answer they get is built from whatever the engine can find. Two of the five dimensions live here. Accuracy scores every factual claim the engine makes about you against the truth: services, locations, prices, people, years trading. Sentiment scores the tone, from a plain recommendation to a caveat, and records where the caveat came from. Invented prices and stale addresses are the commonest findings.

The deliverable

What do you get for the fee?

A branded report, delivered as a document you can share with partners or your board. Seven sections, evidence throughout.

Headline verdict: your blended AI visibility score and the single most important finding, first.
What AI tells your customer: every discovery prompt with a screenshot and whether you appeared.
What it gets wrong: every false or invented claim about your business, engine by engine, with the correct fact and the commercial risk.
Who it recommends instead: the competitors named ahead of you, and on what basis.
Full scorecard: five dimensions across four engines, plus the blended score.
Ranked fix plan: the top five actions, ordered by impact against effort, each tied to a specific finding.
90-day target: what a re-audit should show once the fixes land, so progress is measurable rather than asserted.

Why not a £49 automated scan? Scanner products check your metadata and schema markup, which matters, and the audit checks it too. But a scanner cannot tell you that ChatGPT recommends the firm two streets away when someone asks for the best surveyor in your town, or that an AI has invented a price you do not charge. Those are the findings that cost you enquiries, and finding them takes a person running real questions and reading every answer.

After the audit

AI search visibility services: audit, fixes, re-audit

The report is yours either way. Three routes from there.

Do it yourself

The fix plan is written to be actionable by your own team or web developer. No lock-in, no retainer required.

We deliver the fixes

Structured data, entity consistency, question-shaped content and the corrections that stop AI answers misdescribing you, scoped and priced from the plan. Where the fixes need a system behind them, that is the AI Implementation engagement, and ongoing upkeep sits on its care plan. A re-audit at 90 days uses the identical prompt set to prove movement. Where the fixes also need the enquiries they produce to land somewhere, that is the small business AI CRM.

Conventional SEO, via partners

Where findings call for technical SEO, link building or content at scale, we introduce specialist partners rather than stretching beyond the specialism. You get the right shop for the work; we stay accountable for the introduction.

The visibility score is also one section of the £350 review on the AI Implementation page, so a business that wants the whole picture of where it leaks enquiries can start there instead.

Where

AI visibility in London and across the UK

The audit runs from London, with every prompt set to a UK location and, where it matters, to your town. A London business gets London prompts; a Manchester clinic gets Manchester ones. The 30-minute scoping call and the report walkthrough happen in person across London, the South East and the Midlands, or on a call anywhere in the UK. Nothing about the method changes with distance.

FAQ

Questions businesses ask about AI visibility

What is an AI visibility audit?

An AI visibility audit is a manual test of what AI assistants say when someone asks them to recommend or describe a business like yours. It scores presence, accuracy, sentiment, competitive position and source basis across four engines, and maps every weakness to a fix. It answers a question your analytics cannot: what AI tells your next customer about you.

How is this different from an SEO audit?

An SEO audit measures how you rank in a list of links. This measures whether you exist inside the single answer an AI gives, and what that answer says. A business can rank well on Google and still be invisible, misdescribed or recommended against inside AI answers. The two are related but scored on different evidence.

How much does an AI visibility audit cost in the UK?

This one is £750, with a launch rate of £500 for three audits a month. UK prices range from £49 automated scans to enterprise deep-dives at several thousand pounds. This is a manual audit run by one person, scoped to one business and its local market, with the evidence in the report.

Do I need an AI visibility agency?

Usually not. An AI visibility agency makes sense for a brand with many locations, many product lines or a large content team, where the fixes are a programme rather than a list. For an owner-led business the findings are typically five things: inconsistent facts across listings, a site that never states its prices or places in plain text, missing structured data, no page that answers the questions customers actually ask, and nobody else writing about you. Those are fixable by your own team or a web developer from the ranked plan, and we re-audit at 90 days to prove it. If the plan turns out to need more hands than you have, we will say so and introduce the right partner.

How long does the audit take?

You receive the full report within five working days of confirming the engagement. We need around 30 minutes of your time up front to confirm your locations, services and main competitors, and the ten prompts are agreed with you before they run so the audit tests the questions your customers actually ask.

Can you guarantee we will appear in ChatGPT answers?

No, and you should be wary of anyone who says yes. AI engines change constantly and nobody controls their output. What we can do is measure your position with a repeatable method, fix the causes found, and re-audit with the identical prompt set so any movement is demonstrated rather than claimed.

Which AI tools does the audit cover?

Four engines: ChatGPT with search enabled, Google AI, Perplexity and Claude. Each is queried with the same fixed set of ten customer prompts under controlled conditions, UK location set and a fresh session per prompt, and every answer is screenshotted and scored. Other engines can be added by scope if your customers use them.

Three audits a month

Book a 20-minute call

£500 fixed at the launch rate, report in five working days, no retainer required. The worst case is you learn AI already recommends you, and you have the evidence to prove it.

Or email [email protected] and we reply within one working day.