Find a problem worth solving
Good products start with a clear problem. Today you turn one of your three problems into a short statement, and get ready to ask a real person about it.
What you will leave with
- Why talking about problems beats talking about solutions
- A five-part problem statement
- Five good questions about real past events
- A rehearsal with the AI, then one real conversation
YOU LEAVE WITH · A problem statement backed by one real conversation
Problems before solutions
Solution talk
- “We need an app for the rota”
- Jumps straight to an answer
- Hard to check with anyone
Problem talk
- “Volunteers miss shifts because swaps get lost”
- Describes what actually happens
- Easy to check with a real person
If your sentence names an app or a website, it is a solution. Try again without it.
A clear problem statement in five parts
- WHOA specific group of people, not “everyone”
- WHENThe situation where the problem happens
- STRUGGLEWhat goes wrong for them
- COSTWhat it costs them: time, money, worry
- EVIDENCEHow you know: guessed, heard from someone, or seen yourself
Evidence means how you know. Be honest: “I guessed” is a fine place to start.
Good questions, poor questions
Poor: invites a polite yes
- “Would you use an app for this?”
- “Don’t you think the rota is a pain?”
- “How much would you pay for it?”
Good: asks about real past events
- “Tell me about the last time a shift was missed.”
- “What did you do about it?”
- “What happened next?”
Ask about what people did, not what they might do. People are kind, so future questions get polite yeses.
Riverside Café (fictional): a first problem statement
First try: “The café needs a rota app.”
Better: “Shift leads at Riverside Café (who), when a volunteer swaps a shift at short notice (when), often don’t hear about it in time (struggle), so the café opens short-staffed about once a week (cost). Heard from one shift lead (evidence).”
Ask
- Which parts were missing from the first try?
- What would you ask the shift lead next?
- Who else at the café might see this differently?
Live demo: rehearse with the AI
I want to practise asking someone about a problem. Please play a busy café volunteer. You are invented, not a real person. Answer my questions briefly, the way a real person would, and if I ask a leading question, give a polite yes. When I type “stop”, step out of the role. Tell me which of my questions were leading or about the future, and suggest a better question for each one, about a real past event.
Check the response
- The AI stays in role until you type “stop”
- It spots at least one leading question
- Each weak question gets a better rewrite
- You treat the rehearsal as practice, not evidence
Your tutorial: 35 minutes
Work through the Lesson 2 tutorial. Each step is a short prompt, then a “Did you…?” check.
- Pick one of your three problems
- Ask the AI to help you write it in five parts, using no real names
- Write five good questions, then rehearse them with the AI
- Ask one real person, with their permission, and note what they said and did
A problem statement backed by one real conversation
Before you move on
- I can tell a problem from a solution
- I have a problem statement with all five parts
- I have five questions about real past events
- I have spoken to one real person, or booked a time to
NEXT · Lesson 3: The smallest useful thing
© 2026 MVP Studios Ltd. The Build with AI: First Steps course materials are free to use, print, translate and adapt for non-commercial training under the Creative Commons Attribution-NonCommercial-ShareAlike 4.0 licence: credit MVP Studios Ltd and share adaptations on the same terms. MVPAI™ and the MVP Studios name are trademarks and are not licensed. Course terms