A clear engagement from the start
Start with a defined brand, market and audience journey. Multi-brand, multilingual and international programmes are scoped separately.
Timing and delivery
Delivered as a scoped research project. Collection windows, review dates and handover are agreed after the markets, platforms and sample are defined.
You work directly with Rob alongside your nominated lead and implementation specialists. Your written scope sets the remote meeting schedule, review points and handover, so everyone knows what to expect.
How the work runs
Agree the commercial question
A remote kickoff establishes the audience, buying decisions, products, markets, competitors and what leadership needs to decide. We agree what a useful baseline must answer.
You receive: Scope brief, access requirements and agreed research questions.
Design and capture the baseline
Build a prompt sample across discovery, comparison and validation. Record the platforms, collection conditions and repetition plan before testing, then preserve the observed answers and sources.
You receive: Prompt register, collection protocol and evidence log.
Investigate the gaps
Separate mentions, recommendations, citations and factual accuracy. Examine competitor patterns, relevant external sources and technical barriers affecting important owned pages.
You receive: Findings tied to evidence, with limitations and alternative explanations.
Turn findings into a working plan
Prioritise changes with the people who can deliver them. Walk the team through the findings and establish how it can repeat the measurement after implementation.
You receive: Decision deck, prioritised roadmap and repeat-measurement instructions.
Inside your delivery pack
A baseline you can inspect
An editable prompt register and evidence log with test conditions, answer references, source URLs and a defined scoring rubric.
A clear account of your position
Findings covering brand representation, recommendation patterns, competitor inclusion and recurring cited sources within the agreed sample.
An implementation roadmap
Recommendations grouped by technical, content, brand or external-evidence work, with owners, dependencies, priorities and validation criteria.
A team handover
A findings walkthrough and written measurement instructions. Meeting count, participants and the clarification period are specified in the proposal.
What a useful finding looks like
Illustrative example for a fictional business. This is a demonstration of the delivery format, not a client result.
- Observation
- In the saved answer to a comparison prompt, a competitor is recommended with a cited product guide; the fictional brand is absent.
- Evidence to retain
- Prompt ID, run date, model and search mode, full answer, citation URL and relevant source excerpt.
- Interpretation
- The cited guide answers a buying question. One answer does not establish that creating a similar guide will cause inclusion.
- Next action
- Check whether the question recurs in the sample and whether the brand has accessible, accurate evidence answering it. Assign a content owner if a gap is confirmed.
- Validation
- Verify the published improvement, then rerun the unchanged sample under recorded conditions and report variation alongside any change.
What your team needs to provide
- A marketing or search lead who can make scope decisions.
- Your priority products, audience journeys, markets and competitors.
- Access to relevant analytics, Search Console and site evidence where available; access gaps are recorded.
- Content, technical and brand contacts for validation and roadmap ownership.
Scope and responsibilities
- The baseline measures an agreed sample; it is not a census of every answer an AI system can produce.
- Content production, development, PR placement and continuing monitoring require an additional scope.
- Tool subscriptions, paid data and collection costs are agreed before they are incurred.
- Recommendations and rankings cannot be guaranteed. Changes are evaluated against documented conditions.
We confirm review rounds and the clarification window in the proposal. If requirements change, we agree the effect on fees and delivery dates before additional work begins.
Scope, fee and next steps
A project fee is quoted after a short scope review. Brand and market coverage, languages, platforms, sample size, repeat runs and technical complexity determine the work. The proposal states the deliverables, dates, meetings, revision allowance and any third-party costs before you commit.
Share your website, what you need to decide and your preferred timing. Include an indicative budget if you have one. You will get a reply within one working day to explore the fit, then an agreed scope and fee before any work starts.
What the audit measures
Seeing your brand in an answer does not tell you the whole story. Is the description accurate? Are you recommended for the right need? Are competitors appearing where you are absent? Your audit records mentions, recommendations, citations and accuracy separately so you can see what needs attention.
The prompt sample is designed around real audience decisions. It can include discovery, comparison, validation, troubleshooting and purchase questions, with important variations documented rather than selected after the answers are seen.
A baseline with boundaries
AI outputs can vary by model, time, location, account state and phrasing. The audit records the environment, date, prompt and sampling method. It does not pretend that a limited observation is a permanent universal result.
- Share of tested prompts in which the brand appears
- Recommendation position and competitors mentioned alongside it
- Cited URLs, domains and recurring source types
- Accuracy and consistency of important brand claims
- Topic and journey gaps
- Owned-page technical eligibility and retrievability
- Signals from reviews, reference sources and earned coverage
From observation to action
Findings are grouped by the team able to act: technical, content, digital PR, product, brand or analytics. Recommendations explain the supporting evidence and expected mechanism rather than attaching certainty to an opaque model.
The repeat-measurement specification is part of the deliverable. It defines which prompts should stay stable, what can change, how results will be recorded and which changes are large enough to investigate.
A useful first engagement
You can use the audit as a standalone starting point, brief your existing agency from the findings or build a broader strategy around them. You receive the evidence and repeat-measurement instructions so your team can use the work after handover.
What this helps you achieve
- A defensible starting point for an AI-search programme
- Evidence that leadership and delivery teams can interrogate
- A clear distinction between visibility, sentiment, citation and referral traffic
- A measurement process that can be repeated after changes
Common questions
What is an AI visibility audit?+
It is a documented assessment of how a brand, product or topic appears in relevant AI-assisted searches. A useful audit examines inclusion, recommendations, descriptions, citations, competitors and the sources that may be shaping the result.
Which AI platforms do you test?+
The platform set is agreed around the audience and market. It can include Google AI features and major conversational assistants, with versions, dates and test conditions recorded.
Can the audit be repeated later?+
Yes. Repeatability is designed into the baseline, including the prompt sample, collection method, measures and known limitations.