Nonlinear / Blog / Measurement
How to build a question bank to measure your AI visibility
A question bank is a fixed set of buyer questions you send to AI models every month. Group them into brand, intent and comparison, keep them stable, and report rates.
Why a bank, not a spot check
Asking an AI about your product once tells you almost nothing, because answers vary. A fixed bank of questions, run on a schedule, tells you whether you’re moving.
Three groups
- Brand questions. “What is [your product]?”, “Is [your product] any good?”. These test whether AI knows you at all.
- User-intent questions. The questions buyers ask before they know your name: “best tool for X”, “how do I do Y without Z”. This is where new customers come from, so most of the bank should live here.
- Competitor questions. “[Competitor] alternatives”, “[Competitor] vs [you]”. These show who gets named in comparisons.
Getting the questions
Don’t invent them. Pull the wording from places buyers talk: Reddit threads, forum posts, support tickets and sales calls. Use their phrasing, not yours.
Running it
- Send every question to each major model, on the same schedule.
- Run each question more than once.
- For each answer, record: named, cited, which pages, and who else was named.
Reporting
Report named rate and cited rate by group, and list who’s named instead. Keep the original bank fixed so months are comparable. Add new questions as a separate group, so you don’t break the trend. For what to do with the gaps, see the GEO guide.
FAQ
How many questions do I need?
Enough to see a trend. We use about 50 across three groups. Start with 20 to 30 if you’re new to this, and keep them fixed.
Why run each question more than once?
AI answers vary from run to run. Rates across several runs are more reliable than any single answer.