Chapter 16
Pretend shoppers give you questions to try
Imagine a toy shop that wants to know what children might ask its robot. So it fills a box with cards, and a helper robot writes a pretend shopper on each one.
Your customers can ask robots which business to choose, and you rarely hear what they ask. So someone may offer you invented customers, which researchers call personas, to write test questions.
Daniil Dzemesjuk and I built such a box, with about a million cards. An invented card can suggest a question that was missing from your list.
The short version. A pretend shopper is an invented person on a card, written by a robot. Its question can help you look for a need your list missed.
About a million invented cards
Every card describes an invented person. The cards include questions for learning, comparing and buying.
What is in the box depends on what went in
Picture it. Each small card is 1 of every 100 pretend shoppers. A robot gave every card a label for its kind of business. Green cards carry one of the three biggest: general, education technology and consulting.
At Daniel’s: like 511 folders for kinds of case, with 3 folders holding about 60 of every 100 client files.
The box leans toward education and careers because the collections did, and most descriptions are American. A big box can still be uneven.
One invented card can add a question
Person: A parent looking for an evening course.
Need: Compare courses that fit around work and childcare.
Stage: Comparing.
Question: “Which evening courses let parents study part-time?”
That question can expand a list that only covered price or course subjects. It adds a need: fitting study around family time.
How to check this for your business
Daniel gets a report with invented questions. He picks one question about choosing a lawyer quickly. Then he opens his meeting notes and finds how a customer asked about that need in real life. He compares the two and asks whether the invented question covers a need missing from his report.
Try this on Monday
Pick one invented question from a report. Find one real customer question about the same need, in the customer’s own words.
Put them side by side. Does the invented question cover a need missing from your list? If yes, add that need to the questions you test. Keep the real wording too.
References
Chapter 16. Pretend shoppers give you questions to try
Preprint. Dmitrij Żatuchin and Daniil Dzemesjuk (2026). Demand-Side Measurement for Generative Engine Optimization: Constructing and Validating a Million-Persona, Intent-Annotated Buyer Corpus. arXiv:2608.30023, version 1, 30 August 2026. arXiv:2608.30023.
Public sample. PersonaGen-15K, 14,955 pretend shoppers. huggingface.co/datasets/rankfor/PersonaGen-15K. Research use only, CC BY-NC-SA 4.0.
What does AI say about your business?
Put the book into practice. Tell us your company and website to request an AI visibility audit. We’ll reply by email to agree what to check.
What the audit shows
- Where your company appears in the answers, and next to whom
- The pages AI cites in your market
- What to publish first on your own site
