Which prompts should you track in an AI visibility report?

The prompts in an AI visibility report should represent questions that matter to the business.

You do not need to track every possible way someone could mention your brand. Start with the buyer questions you care about, run them consistently across the AI engines you want to monitor, and keep the set stable enough to compare results over time.

A useful starting set usually includes brand, category, comparison, and use-case prompts.

Track prompts that represent real buyer questions

Start with the decisions a potential customer is trying to make.

For a B2B software company, that might include:

• understanding what the product does
• finding tools in a category
• comparing the brand with an alternative
• checking whether the product fits a specific use case
• confirming an important capability

Then turn those decisions into natural questions.

For example:

Prompt type	Example
Brand	What is Northstar CRM?
Category	What are the best CRMs for small agencies?
Comparison	Northstar CRM vs Orbit CRM: which is better for agencies?
Use case	Which CRM is best for a five-person agency that needs client reporting?

You do not need an equal number from every type. Track the questions that matter for the brand and audience you are reporting on.

Include brand prompts

Brand prompts show how AI systems describe the company directly.

Examples:

• What is [brand]?
• What does [brand] do?
• Who is [brand] for?
• What are the main features of [brand]?

These can reveal whether an AI answer understands the basic product, audience, and positioning correctly.

They are especially useful when the brand has changed positioning, launched a new product, or is frequently described inaccurately.

Include category prompts

Category prompts show whether the brand appears when someone asks for products in the market.

Examples:

• What are the best [category] tools?
• What are the best [category] tools for [audience]?
• Which [category] platforms should a small business consider?

These are useful for measuring whether the brand appears alongside the competitors you expect.

Avoid category questions that are too broad to be useful. "What are the best software tools?" is unlikely to tell a CRM company much.

Include comparison prompts

Comparison prompts represent buyers deciding between products.

Examples:

• [Brand] vs [competitor]
• What are the best alternatives to [competitor]?
• How does [brand] compare with [competitor]?
• Which is better for [audience], [brand] or [competitor]?

These prompts can show whether your brand appears in the competitive set and how AI systems describe the differences.

Track the competitors that actually matter to the business rather than building a long list just because they exist.

Include use-case prompts

Use-case prompts connect the product to a specific customer need.

Examples:

• Which [category] tool is best for [audience]?
• What is the best [category] tool for [requirement]?
• Which product supports [important use case]?
• What should a [type of company] use for [job]?

These can be more useful than broad category questions because they test whether AI systems understand where the product fits.

For example:

Which CRM is best for a five-person agency that needs client reporting?

That tells you more about positioning than:

What are the best CRMs?

Start with a small prompt set

You do not need hundreds of prompts to start tracking AI visibility.

Five to ten important buyer questions can give you a useful first view.

For example:

1. What is Northstar CRM?
2. What are the best CRMs for small agencies?
3. What are good alternatives to Orbit CRM?
4. Northstar CRM vs Orbit CRM for agencies
5. Which CRM is best for a five-person agency?
6. Which CRM has good client reporting?
7. What CRM should a growing agency use?

Use more prompts when you need to cover additional products, audiences, markets, competitors, or buyer decisions.

The goal is not to maximize the number of prompts. It is to track questions where the answer would actually matter.

Keep a simple prompt worksheet

Record enough information to understand why each prompt exists.

A simple worksheet can include:

Field	Example
Prompt	Which CRM is best for a five-person agency?
Type	Use case
Audience	Small agencies
Competitors	Northstar CRM, Orbit CRM
Why we track it	Agency fit is an important sales question
Added	20 August 2026

You can record the AI engines and reporting cadence at the report level rather than repeating them for every prompt.

The worksheet is there to keep the prompt set understandable, not to turn prompt tracking into a governance exercise.

Run the same prompts over time

Once you choose the set, keep the wording stable when you want to compare results.

Run the prompts across the AI engines you track and record:

• whether your brand appeared
• which competitors appeared
• how the brand was described
• which sources were cited when relevant
• the date of the run

Then repeat the same set later.

Changing:

What are the best CRMs for small agencies?

to:

What CRM should a marketing agency with ten employees use?

may give you a useful new question, but it is not the same measurement.

Add it as a new prompt rather than silently replacing the old one.

Add prompts when the business gives you a reason

Your prompt set should evolve as the business changes.

Add a prompt when:

• sales repeatedly hears a new buyer question
• a new competitor becomes important
• the company enters a new market
• a new product or feature changes positioning
• a high-value use case becomes a priority
• an AI answer reveals a question worth tracking separately

Remove prompts when the underlying buyer question is no longer relevant.

Record the change so you know when two reporting periods are no longer directly comparable.

Track prompts with Maverank

You can manage a small prompt set manually in a spreadsheet.

For ongoing reporting, Maverank lets you track the buyer questions that matter across ChatGPT, Claude, Gemini, and Google AI.

For each prompt, you can review whether your brand appears, which competitors are mentioned, how visibility differs by AI platform, and how those results change over time.

That makes it easier to keep the same prompt set across reporting periods and investigate the questions where your brand gains or loses visibility.

It also connects prompt tracking with broader measurements such as AI share of voice, so you can move from a headline visibility metric to the individual buyer questions behind it.

Start with a small set of important prompts, then add new ones when another audience, competitor, product, or use case becomes worth tracking.

Frequently asked questions

How many prompts should an AI visibility report track?

Start with five to ten important buyer questions. Add more when you need to cover another product, audience, competitor, market, or buyer decision.

What types of prompts should I track?

Brand, category, comparison, and use-case prompts are a useful starting point. You do not need to use all four if one type is not relevant to the business.

Should I track the same prompts on every AI engine?

Yes, when you want to compare visibility across engines. Run the same wording on each platform and record failed or missing responses.

Should I use search volume to choose AI prompts?

Search data can help identify topics people care about, but traditional search volume is not the same as AI prompt demand. Use it as one input alongside sales questions, customer research, positioning, and business priorities.

When should I change the prompt set?

Change it when the business changes: a new audience, competitor, product, market, or buyer question becomes important. Record the change so you know when two reports are no longer directly comparable.

Can Maverank track these prompts?

Yes. Maverank lets you track selected prompts across ChatGPT, Claude, Gemini, and Google AI and review how your brand and competitors appear over time.

Start with the questions that matter

Choose a small set of questions that represent how buyers discover, compare, and evaluate your brand.

Track them consistently and expand the set when another buyer question becomes important.

Try Maverank to track those prompts across ChatGPT, Claude, Gemini, and Google AI.