Product manager
AI Prompts for User Stories
Start with Step 1See an example
- Who it is for
- Product managers, product owners and founders who need to turn a feature idea into stories the team can estimate and build, without losing the reason behind it.
- Outcome
- A small set of user stories built from the problem, users and notes you supply, each with acceptance criteria, edge cases, open questions and a suggested split if a story is too big. You review and refine them with your team.
These prompts are plain text, so they work in whichever AI chat tool you use. Paste each step into the same chat so the AI keeps the context from earlier steps.
Step-by-step prompts
How to use it: replace the words in [brackets] with your details, paste Step 1 into your AI tool and answer any questions it asks, then paste the next steps into the same chat. Read and edit the result before you use it.
Step 1
Describe the problem first
I'm writing user stories for a feature. Here's the context (I've removed confidential or customer-identifying details): - The problem we're solving: [DESCRIBE] - Who has the problem: [USER TYPES] - What we've learned from users so far: [PASTE NOTES OR SUMMARIES] - What's out of scope: [LIST] Use only this context. Before writing stories, restate the problem in one sentence and list any assumptions you'd be making.
Step 2
Draft the stories
Write 3–6 user stories in the format "As a [user type], I want [goal], so that [reason]". Each story should describe something a user can do or see, not a technical task. Keep the "so that" tied to the problem from Step 1. If a story isn't supported by the context, mark it "ASSUMPTION".
Step 3
Add acceptance criteria
For each story, write 3–5 acceptance criteria in "Given… When… Then…" form. Make each one checkable by someone testing the feature by hand. Don't specify the design or technology unless I gave it.
Step 4
Find edge cases and questions
For each story, list: - 2–3 edge cases (empty states, errors, permissions, unusual inputs) - Any open question for design, engineering or the people who requested it Then point out any story that's too big to finish in one iteration and suggest how to split it into smaller stories that each still deliver something to the user.
Step 5
Check the set before refinement
Review the full set. Check that each story is independent, valuable to a user, small enough to estimate and checkable. List any that fail and why, and give me the revised set ready to paste into our backlog.
Example of good output
Illustrative example, fictional product; not real user research or model output:
Story: As a team admin, I want to invite a teammate by email, so that they can join our workspace without asking support.
Acceptance criteria:
- Given I'm an admin, when I enter a valid email and select Invite, then the teammate receives an invitation email.
- Given the email already belongs to a member, when I select Invite, then I see a message saying they're already in the workspace.
- Given I'm not an admin, when I open the members page, then I don't see the Invite option.
Edge case: The invitation link has expired.
Open question: How long should an invitation stay valid?
Tips and pitfalls
- Start from the problem, not the solution. Stories that begin with a feature idea tend to skip why it matters. The "so that" line keeps the reason visible.
- Mark assumptions clearly. Anything not backed by research or a stakeholder request should be labeled so the team can check it.
- Write criteria a person can check. If nobody can tell whether a criterion is met, rewrite it.
- Refine with the team. Treat the output as a first draft for your refinement session. Engineers and designers will find gaps the AI can't.
- Paste only what's needed. Use an AI tool your organization allows, and leave out customer names, contracts and internal financials.
FAQ
Can I paste interview notes or feedback directly?
Yes, if your company allows it. Remove names and identifying details first, and paste summaries rather than full transcripts where you can.
What if my team uses a different story format?
Paste your team's format into Step 2 and ask the AI to follow it.
Does this replace a product requirements document?
No. Stories describe what users can do. A requirements document covers goals, scope, success measures and decisions, and you can paste it into Step 1 as context.