1. Start with a customer need
A useful question connects something the customer needs with something the client offers. Begin with the services and customers in the reviewed business profile, then ask the client which commercial priorities matter most.
For fictional Dallas HVAC business ClearAir, emergency repair, heat-pump installation, and maintenance represent different needs. If the agency focuses only on emergency repair, the results will say little about installation visibility. Choose the scope deliberately.
Top Mind helps generate questions based on the business. Your agency reviews and edits them before use. Treat the generated set as a draft that benefits from client knowledge.
2. Cover different buying situations
Use the examples below as starting points, adapting services and locations to the client’s actual coverage. These questions are illustrative; they are not search-volume estimates or evidence of how frequently customers ask them.
| Customer intent | Example question | What to investigate |
|---|---|---|
| Find help now | Who offers emergency AC repair in Dallas? | Whether ClearAir appears for an urgent service need. |
| Find a specialist | Which HVAC companies in Plano install heat pumps? | Whether relevant providers, including the client, are named. |
| Evaluate an option | What should I look for when choosing an AC installation company? | Which decision criteria and sources the answer emphasizes. A brand mention may not be necessary to answer this question. |
| Compare providers | How should I compare quotes from AC installation companies in Dallas? | What information customers may need when evaluating offers. |
| Research a brand | What services does ClearAir in Dallas offer? | How AI describes the named business. |
3. Separate discovery from named-brand research
Unbranded questions do not supply the client’s name. They help investigate whether AI brings the business into a relevant conversation on its own. “Who offers emergency AC repair in Dallas?” belongs in this group.
Branded questions include the client’s name. They help you review how AI describes a known business, including its services or positioning. A mention in this group does not establish that AI would independently recommend the brand.
Distinguish the two groups when reviewing your questions in Top Mind and presenting the results. They answer different questions and should not be combined into a broad claim of spontaneous discovery.
4. Edit for relevance and clarity
- Check the service. Is this an offering the client actually provides and wants to discuss?
- Check the location. Does the business serve that area? Do not add nearby cities simply to expand the list.
- Use natural wording. Write a question someone could plausibly ask during a buying decision.
- Avoid supplying the desired answer. A question describing the client as “the best” cannot tell you whether AI would reach that conclusion without your suggestion.
- Know what you will inspect. Some questions concern provider discovery; others concern descriptions, criteria, or cited sources.
For ClearAir, review Dallas and Plano questions with the client before keeping both locations. A geographically specific question is useful only if that market matters to the business.
5. Keep a baseline and make changes explicit
After reviewing your starting set, keep the reviewed questions in Top Mind as your baseline. Use a consistent set for repeat scans so that changes in questions are not confused with changes in visibility.
You can still explore new questions. Label them as additions and explain why they matter. If you rewrite a question substantially, describe the new scope instead of presenting its result as a direct comparison with the old wording.
When reporting, include the questions, providers, dates, and relevant settings where available. Repeated scans help provide context, but consistent wording does not make AI answers deterministic.
Next, use the assessment guide to inspect answers, or the reporting guide to explain changes to clients.