How do I find the questions my customers are asking AI?

Arjita Sethi · Updated November 11, 2026 · 4 min read

AI for Growth

You find the questions your customers are asking AI by collecting the questions they have already asked you and others, in reviews, DMs, emails, sales calls and forums, and then sorting them by how close the person is to buying. You don't need to guess or invent them. Once they are sorted, the questions where people compare options or decide between businesses are the ones to answer first, because those are the questions where ChatGPT, Claude and Google's AI answers actually name businesses. Build with AI Club teaches this as the Question Map.

Why the questions matter more than keywords now

People used to type two or three words into a search box. Now they type a full sentence into an AI tool, and they include who they are and where they live. "Best bookkeeper for creative agencies in Oakland" is a question a small business can win. "Bookkeeper" goes to the famous names every time.

That shift means the exact words your buyer uses are now the most valuable research you have.

Step 1: Pick one buyer

A question map only works when it is built around one buyer. In class I use a fictional sample business, Juniper Books, an Oakland bookstore run by Dana. Dana picked debut authors in the East Bay who are planning a book launch, because a launch night is her highest-value booking. She still sells to everyone who walks in, and she leads with one buyer when she speaks.

Write your buyer as one sentence that includes who they are, where they are and the moment that makes them ask for help now. For Dana, the moment is a set publication date, a publisher who won't fund a tour, and friends asking where the party is.

Step 2: Collect questions from four places

  1. Reviews, your own and the ones on businesses like yours.
  2. DMs and emails, the questions people actually sent you.
  3. Sales call notes, especially objections, written in the buyer's words.
  4. FAQs and forums like Reddit, where people ask strangers what they would never ask you directly.

Copy the questions word for word. The way a buyer phrases a question is part of the data.

Step 3: Sort the questions by mood

Paste everything into your AI project and ask it to sort each question into one of three moods.

  • Exploring: "How do book launches work?" These need an explainer.
  • Comparing: "Best East Bay stores for a reading?"
  • Deciding: "Is Juniper Books good for a launch?"

Comparing and deciding questions are where AI names businesses, so that is where you choose your must-win questions.

Step 4: Pick ten must-win questions and find the gap

Choose ten questions you would most like to be the answer to. Three of Dana's are "Where can a debut author hold a book launch in Oakland?", "Do indie bookstores charge to host a reading?" and "How do I get 40 people to my first book event?"

Then look for the gap, the question nobody nearby answers well. Dana found that no store near her explains what a launch night costs or includes. That gap is the first page she writes.

FAQ

How many questions should I collect? Collect as many real ones as you can find, then narrow to ten must-win questions. Real questions from real buyers beat a long list you made up.

Can AI find the questions for me? AI is very good at sorting and grouping the questions you collect. Start with your own inbox and calls, because AI can't see your DMs or sales calls unless you give them to it.

What do I do with the ten questions? They become the brief for your website pages, your articles and your posts, and they are also how you check later whether AI recommends you.

Learn it with me

This week's Build with AI Club YouTube lesson, 4 Claude skills I would set up before anything else, shows how to save work like this as a skill, so you don't explain your buyer to AI twice. Watch it on the Build with AI Club YouTube channel

I'm teaching AI 101 at SHACK15 in San Francisco tonight, Wednesday, Nov 11, 6 to 7 PM PT. It's Your AI Research Desk, and free online audit seats are open at luma.com/build-4bhm. Tomorrow, Thursday, Nov 12, 10 to 11 AM PT, Sarah Dooley and I host Aligning AI With Your Family's Values, free on Zoom, at luma.com/build-nt03.