AI Visibility Audit
Understand what AI says about you — and what evidence it has to work with.
See how selected AI systems represent your organisation across the questions buyers actually use to compare, shortlist and decide.
The AI Visibility Audit connects that representation back to your evidence, positioning, customer proof and wider information environment, so you can distinguish an optimisation problem from a knowledge problem.
AI visibility is not only a search problem
Organisations are beginning to ask why competitors appear in ChatGPT, Perplexity, Google AI experiences and other AI-generated answers while they do not.
The temptation is to respond with more prompt tracking, more pages, more FAQs and more optimisation. Those can be useful. They cannot compensate for an organisation whose positioning is vague, whose evidence is fragmented or whose expertise is difficult to verify.
There is another complication: some buyers may use an AI assistant to orient themselves before they ever reach your website. By the time they arrive, they may already have a shortlist and a set of decision criteria. The audit therefore starts with buyer questions rather than a generic prompt list.
- Important buyer questions are not clearly answered anywhere.
- Product positioning is too vague or inconsistent to survive synthesis.
- Valuable research and customer proof exist but are scattered across unrelated pages.
- Third-party sources describe the organisation differently from its own website.
- Teams can measure referral clicks but not the AI-assisted journeys that never preserve a referral path.
The audit identifies the editorial, evidence and measurement problems beneath the visibility problem.
What you leave with
Seven concrete deliverables, shaped around your market, buyer questions and available evidence.
Buyer-question and decision-criterion map
A focused map of the questions, requirements, comparisons and constraints that define the buying territory being tested.
AI representation baseline
A documented view of how selected AI systems currently represent the organisation and relevant competitors across agreed scenarios.
Evidence and authority map
A map of the research, expertise, customer proof, product facts and external signals currently supporting the organisation's authority.
Positioning and decision-support diagnosis
An assessment of whether differentiation, use cases, proof and buyer reassurance are explicit enough for people and AI systems to interpret.
Evidence-gap and signal-gathering plan
A prioritised list of questions the organisation cannot yet answer strongly enough, with the most credible routes to better evidence.
Measurement plan
A three-layer model covering AI representation, observable traffic and conversion, and qualitative reconstruction of how buyers actually searched.
Prioritised 90-day roadmap
A practical sequence of technical, editorial, evidence and positioning changes, including what should be re-tested.
The exact scope is agreed before the audit and shaped around your market, priority questions, competitors and available evidence.
What the audit assesses
Six connected areas that show how the organisation is represented, what supports that representation and where the learning loop is weak.
Buyer questions
- Real customer language
- Decision criteria and constraints
- Use-case differences
- Recurring objections and risks
- Comparison logic
- Priority commercial questions
AI representation
- Presence across agreed scenarios
- Accuracy of company descriptions
- Topic and category association
- Citations and source patterns
- Competitor representation
- Missing, vague or wrong claims
Evidence strength
- Original research
- Customer evidence and case studies
- Product facts and proprietary insight
- Named experts
- External corroboration
- Clear sourcing and provenance
Positioning and decision support
- Clear differentiation
- Explicit audiences and use cases
- Comparable customer proof
- Implementation and risk evidence
- Commercial claim support
- Consistency across product pages
Distributed reputation
- Owned and third-party consistency
- Customer reviews and references
- Media and analyst signals
- Expertise across channels
- Source-worthy evidence
- Contradictory public descriptions
Measurement and learning
- AI representation tracking
- LLM referral visibility
- Landing-page and conversion signals
- Discovery-call research paths
- CRM attribution capture
- Re-testing after evidence changes
The assessment applies the AI Visibility and Buyability framework.
Typical engagement
A focused audit covering an agreed set of commercial topics, buyer questions, competitors and available evidence.
Typically delivered over two to three weeks, with a written report, evidence map and findings session.
The scope and fixed fee are agreed in advance. The final report brings together the buyer-question map, AI representation baseline, evidence and positioning findings, signal-gathering priorities, measurement plan and 90-day roadmap.
How the audit works
Define the buying territory
Agree the priority audiences, commercial topics and competitors, then map the buyer questions, constraints and decision criteria that matter.
Test current representation
Sample selected AI systems across bounded scenarios and record recurring descriptions, omissions, sources and competitor patterns.
Audit the evidence and positioning
Assess the organisation's product facts, research, customer proof, expertise, third-party signals and decision-support material.
Design the learning loop
Identify what new evidence is needed and how representation, analytics and buyer-research paths should be measured together.
Set the roadmap
Prioritise the technical, editorial, positioning and evidence work for the following 90 days and define what should be re-tested.
Who this is for
B2B organisations whose credibility depends on being understood in complex or competitive markets.
- SaaS and enterprise technology companies
- Fintech and financial services organisations
- Professional services firms
- Research-led organisations
- Companies investing in thought leadership
- Startups and scaleups entering established categories
It is particularly useful when you already have substantial content or evidence but cannot see it translating into authority, visibility or differentiation.
The work is typically commissioned by marketing, content, communications, research, product marketing or founder-led teams.
What this audit is not
- Not a guarantee of rankings, citations or AI inclusion. AI-generated answers vary by system, prompt, geography and time. The audit improves the underlying information environment — it cannot guarantee a fixed position.
- Not a new keyword-tracking service. Prompts are used as samples of buyer questions and decision scenarios, not treated as a stable ranking system or a replacement keyword list.
- Not only a technical SEO audit. Technical accessibility matters, but the service also examines evidence quality, positioning, expertise, customer proof and third-party reputation.
- Not automatically a recommendation to publish more. The answer may be to connect, strengthen or better evidence what already exists — or to learn something the organisation does not yet know.
- Not based on the assumption that every buyer starts in an LLM. The audit tests AI-mediated discovery as one part of a wider buying journey and keeps the limits of observable attribution explicit.
Why Editorial Intelligence
AI visibility exposes a wider editorial question: what public account of the organisation can a buyer or machine construct from the evidence that exists?
Editorial Intelligence provides the underlying method for answering it: capture useful signals, identify missing evidence, interpret what matters, develop defensible positions and connect those positions across owned and third-party sources.
The audit applies that method to a bounded commercial problem and then closes the loop by asking what the organisation needs to learn, prove or make clearer next.
Keep in touch with Editorial Intelligence
Occasional updates on new research, findings and ways to take part.
Almost there — check your inbox.
A confirmation email is on its way. Your address is only added to the list once you click the link in it, so if it does not arrive, nothing has been signed up.
Founding audit engagements
I am opening a small number of founding engagements to apply and refine the audit with B2B organisations facing real AI visibility questions.
Founding clients receive the complete audit, findings session and prioritised roadmap. I will also ask for candid feedback and, where appropriate, permission to document the engagement as an anonymised or attributed case study.
Frequently asked questions
Which AI systems do you assess?
The exact mix is agreed according to the organisation's market and priority audiences. It may include ChatGPT, Google's AI experiences and Perplexity, depending on where buyers are most likely to encounter AI-generated answers. The audit does not rely on access to private model data.
How do you choose which prompts to test?
The starting point is buyer language and decision context, not a generic list of high-volume prompts. Where possible, scenarios are derived from sales calls, customer interviews, reviews, support conversations, research and the actual requirements buyers use to narrow a shortlist.
Can you guarantee that we will appear in AI answers?
No. AI outputs vary by system, prompt, geography and time. The audit improves the underlying evidence, positioning and accessibility that make an organisation easier to understand and verify. It cannot guarantee a fixed position.
Is this the same as SEO?
It overlaps with search in some areas — content architecture, technical accessibility and topical coverage all matter. But this service also examines evidence quality, product positioning, customer proof, expertise signals and the wider public account of the organisation.
Do we need a large content library?
No. A smaller body of original, well-connected and credible material can be more useful than a large volume of generic content. The audit assesses what you have, what is missing and what needs to be easier to find or verify.
How do you measure AI-assisted journeys that do not click through?
You cannot reconstruct every journey from analytics. The measurement plan therefore combines sampled AI representation, observable referral and conversion data, and qualitative questions in discovery or win/loss conversations about how buyers searched, what they asked and what came up.
Will you implement the recommendations?
Implementation support can be scoped separately — evidence generation, narrative development, content architecture or editorial-system work. The audit is designed to remain useful as a standalone decision product.
Find out what AI can — and cannot — confidently say about your organisation.
Map the buyer questions that matter, establish your current representation and identify the evidence, positioning and measurement gaps behind it.