AI visibility · Brand mentions · Source analysis
Is your business
in the answer?
Find out.
Understand how your business appears when potential customers ask AI tools about your services. KOAT analyses observed brand mentions, linked sources and factual accuracy, then turns the findings into a practical improvement plan.
How do you check whether your business appears in AI answers?
Test a defined set of customer questions across selected platforms. Save the exact questions, answers, dates and test settings. Record brand mentions and source links separately, verify business claims and repeat the checks to understand variation.
What the audit investigates
Understand how your
brand is represented.
An appearance can take several forms. We examine what was actually shown and whether it helps a potential customer understand your business.
Is your business named?
Record whether the answer mentions your business and in what context: a provider list, comparison, explanation or response to a direct brand question.
Is your website cited?
Inspect visible citations and linked pages. Your website may be cited without your brand being named, while a brand mention may appear without a link to your website.
Are the details correct?
Compare claims about your services, location, contact details and market coverage with verified business information. Document inaccuracies for further investigation.
Which alternatives appear?
Record competing businesses and the visible sources associated with them. Look for recurring patterns across the agreed questions and repeated tests.
Does your website answer the question?
Review whether priority pages explain the services, suitability, process and practical details that customers are asking about.
What happens after the click?
Check whether cited landing pages provide an accurate, useful next step. A relevant visit should lead to a clear service explanation and an accessible enquiry option.
Start with search intent
Test the questions
your customers ask.
We build a question set around your services, customer problems, markets and buying decisions. Direct brand questions and unbranded discovery questions are analysed separately.
Asking an AI tool about your company by name tests a different situation from asking it to suggest a provider. Both can reveal useful information, but they need different interpretation.
| Question group | Example | What it investigates |
|---|---|---|
| Problem exploration | “Why is my business website generating few enquiries?” | Which explanations and information sources appear. |
| Service discovery | “Which agencies offer WordPress development in Suriname?” | Observed provider mentions for a relevant market. |
| Provider evaluation | “What should I check before hiring an SEO agency?” | Decision criteria and cited guidance. |
| Direct brand research | “What services does [business name] offer?” | Accuracy and completeness of brand information. |
| Practical buying questions | “What information should I prepare for a website quote?” | Useful content opportunities around the enquiry process. |
These examples illustrate the method. Your audit uses questions selected for your own business and audience.
Documented observations
A useful audit
needs inspectable evidence.
Each recorded observation should be traceable to a specific test. That makes it possible to review the finding, compare later checks and understand the limits of the sample.
| Field | What we record |
|---|---|
| Question | The exact prompt, intent group and language. |
| Test environment | Platform, visible mode or model, search status where observable, session context and relevant location settings. |
| Date and repetition | The test date and its place in the agreed repeat schedule. |
| Answer evidence | A saved response or screenshot showing the observation. |
| Brand presence | Whether the business was named and how it was described. |
| Visible sources | Cited URLs, source types and relevant linked pages. |
| Verification | Checked claims, discrepancies and unresolved questions. |
| Recommended action | A specific next step with evidence and priority. |
What source links can tell you
A visible citation provides a source to inspect. It does not reveal every input used to produce the answer or prove why a particular business was selected.
Interpret the results carefully
Separate a useful sample
from a broad claim.
Repeated checks reveal variation
Answers may differ across platforms, sessions, wording and dates. We repeat an agreed set of checks and report the conditions alongside the results.
If your business is absent from the tested responses, the finding applies to that sample. It does not establish that your business never appears in AI answers.
Define every metric
A sample mention rate is the number of valid recorded responses naming your business divided by the total valid responses in that defined group.
Citation rates are measured separately. These are audit metrics for the tested sample, rather than estimates of platform-wide audience reach.
Example: 6 mentions in 20 valid tests
That is a 30% observed mention rate for those tests. The report should also show the questions, dates and platforms so the figure has a clear meaning.
From findings to priorities
Turn observations
into practical improvements.
Clarify business information
Make your services, market coverage and contact information easy to understand. Investigate inconsistencies across your website and relevant public business profiles.
Improve useful content
Address important customer questions with accurate explanations, suitable examples and verified evidence. Assign an owner to keep priority information current.
Investigate access issues
Review relevant crawl, indexing and page-access problems where the evidence warrants it. Technical accessibility supports discovery but does not guarantee selection.
Recommendations distinguish observed issues from hypotheses. Each action should identify the affected page or profile, the proposed change and how it will be checked.
Our approach
Define. Test.
Analyse. Review.
-
Define the scope
Agree on services, markets, languages, platforms, question groups and verified business facts.
-
Record the tests
Run the agreed checks, capture responses and source links, and document relevant test conditions.
-
Analyse the evidence
Review mentions, accuracy, recurring sources and gaps. Prepare a prioritised action plan.
-
Review progress
Repeat agreed checks after implementation and compare observations with the documented baseline.
Scope and investment
What does an AI
visibility audit cost?
The scope depends on the number of questions, platforms, languages, markets and repeat checks, together with the depth of source and website analysis.
We agree on the sample, reporting format and deliverables before preparing a quote. Implementation and ongoing monitoring can be scoped separately.
What helps us get started?
- Your website and official business name.
- Your priority services and customer groups.
- The markets and languages you want to investigate.
- Verified service and company information.
- Relevant competitors and customer questions.
- Existing AI responses you want us to examine.
Frequently asked questions
AI visibility,
explained clearly.
How do I check whether my business appears in AI answers?
Choose relevant customer questions, run documented tests, inspect mentions and source links, and verify the details. Repeat the checks before drawing wider conclusions.
Which platforms can the audit cover?
The scope can include agreed tools such as ChatGPT, Gemini, Perplexity and available Google AI search experiences. Availability, access and test modes are confirmed when defining the audit.
Is a brand mention the same as a citation?
A mention names your business. A citation links to a source. We record them separately because either can appear without the other.
Does a citation mean the answer is accurate?
Inspect the linked page and compare it with the associated claim. A source link alone does not establish that every statement in an answer is supported.
Why do we test questions without the brand name?
Unbranded questions investigate discovery during customer research. Direct brand questions investigate what an AI tool says when someone already knows your business.
Can one test prove that my business is invisible?
One test records one response under particular conditions. A useful assessment examines a defined set of questions and repeats checks to understand variation.
Can you guarantee that AI tools will recommend my business?
No. We can investigate observed results and improve the information your business publishes. Selection remains outside our control.
Does an AI visibility audit replace an SEO audit?
They address related questions. An SEO audit examines your website’s search foundations, while this audit focuses on observed AI answers, brand representation and visible sources.
Can the audit cover Dutch and English?
Yes. We define separate question groups and interpret their results according to the intended audience and market.
How often should I repeat the analysis?
Set a schedule that matches your priorities, budget and pace of change. Keep a consistent core question set and document any changes to the testing method.
