Practical guide
How to check what AI says about your business
Ask several AI tools the same set of questions your buyers ask, record the answers, and check them for accuracy and mentions of your business.
The work to plan together
A scope shaped around this requirement.
- List 10–20 real buyer questionsScope
- Ask ChatGPT, Gemini, Perplexity and CopilotReview
- Record mentions and errorsCheck
The useful starting point
Create a short question set covering brand facts, service suitability and comparisons. Record the tool, interface, date, question, answer and links. Repeat under reasonably consistent conditions and keep errors separate from absence.
What to review
- List 10–20 real buyer questions
- Ask ChatGPT, Gemini, Perplexity and Copilot
- Record mentions and errors
- Repeat monthly
A practical sequence to follow
Build a useful question set
Collect real buyer questions about suitability, services and comparisons. Separate branded questions from unbranded research. A question designed only to coax a brand mention does not represent the decision a customer is actually making.
Record the checking conditions
Name the tool and interface, question wording, date and any relevant country or language context. Save the answer and cited links. A screenshot without those conditions is difficult to compare with the next check.
Classify what appeared
Distinguish a correct description, a brand mention and a citation to your page. Record factual errors separately. Those are different observations, and none alone proves the answer produced a visit or a qualified enquiry.
Correct sources and repeat
Update your own verified information and request corrections to important external sources when needed. Repeat comparable questions over time and explain variation. A correction to a source does not guarantee an immediate change in the generated answer.
What to avoid or interpret carefully
Results can vary. Do not present a handful of sampled answers as a complete market share or combine different tools without naming them.
Practical detail
A repeatable observation sheet
Use a small buyer-question set and preserve the question wording. The example record below is a method, not a visibility result.
| Part of the work | Example output | How to review it |
|---|---|---|
| Question context | “What does [real business] offer?” plus platform and date | Record the experience and available location/session context. |
| Answer evidence | Brand mention, description and linked source | Separate factual accuracy from a favourable recommendation. |
| Review action | Outdated service statement traced to an old page | Correct the approved source and record a later check. |
Include unsuccessful observations in the same log. Changing prompts until the brand appears can make the result look better without measuring the original question.
Keep an answer record that another person can inspect
A useful check records more than a screenshot saying the brand appeared. Name the tool, interface, date, question and market context. Then evaluate factual accuracy, mentions and citations separately.
Choose relevant questions
Include service suitability, comparison and factual business questions. Avoid a set composed only of prompts that explicitly request the brand. Keep the selection method and negative findings visible.
Inspect the facts
Check services, location, contact information and qualifications. Note which sources appear where available. Incorrect information may come from a stale public record rather than your latest page.
Choose a correction
Update sources you control and use appropriate external correction or feedback routes. Record the action and repeat comparable checks later. Immediate answer replacement cannot be guaranteed.
How to interpret this guidance
A favourable answer is not a confirmed customer outcome. Keep sampled visibility separate from website visits and accepted enquiries.
How often should we repeat the checks?
Choose a practical routine based on the importance of the questions and source changes. Keep the method consistent and document changes. More checks do not automatically create a reliable trend if the sample or interface keeps changing.
How often should we repeat the questions?
Agree a review cadence that fits the purpose and budget. Use consistent conditions and label the findings as sampled observations.
Can Tangensys help with the next step?
Yes. Share your website, goal and the issue you are investigating. We can discuss the relevant service and provide a custom quote within 2 working days.
Related reading and services
Let’s agree the right next step.
Tell us your goals and current setup. We’ll recommend a practical starting point and provide a custom quote.
Custom quote within 2 working days