Turn a Feature Idea into User Stories with Given/When/Then Criteria

Get a backlog-ready set of single-outcome user stories with testable Given/When/Then acceptance criteria, size estimates, and flags on stories too big for one sprint.

You getSummary table and for each story

Prompt

Prompt
Act as a senior product manager who writes backlog-ready user stories for agile teams.

I will paste a feature idea right after this prompt. If it does not say who the users are, what problem it solves, or how long our sprints are, ask me up to 3 short numbered questions in one message and wait for my answers before writing anything.

Then break the feature into user stories.

Rules:
1. Use "As a..., I want..., so that..." with exactly one user type and one outcome. If an outcome contains "and", split the story.
2. Give each story 2-4 acceptance criteria in Given/When/Then form, each with a clear pass or fail result.
3. At least one criterion per story covers an error, empty state, or edge case.
4. Size each story S, M, L or XL. Mark any XL, or anything that would not fit one sprint, as "Too large" and suggest how to split it.
5. Order stories so the first one delivers a thin, usable slice on its own.

Output format:
- Summary table: ID | Story | Size | Sprint fit (Yes / Too large)
- For each story: ID and short title, the story, numbered Given/When/Then criteria, dependencies, split suggestion if flagged.
- Open questions: up to 5.

End with one line listing the assumptions you made and anything missing that could change scope.

Example output

At a glanceSaved search user stories: size and sprint fitFour stories for saved searches with email alerts, 2-week sprints
ItemSizeSprint fit
US-1 Save a searchSYes
US-2 View and delete saved searchesSYes
US-3 Email alerts for new matchesXLToo large
US-4 One-click unsubscribeMYes

Example for: saved searches with email alerts in a rental listings app, renters as users, 2-week sprints.

Summary

IDStorySizeSprint fit
US-1Save a searchSYes
US-2View and delete saved searchesSYes
US-3Get email alerts for new matchesXLToo large
US-4Unsubscribe from alerts in one clickMYes

US-1: Save a search

Story: As a renter, I want to save my current filters, so that I can rerun the search without re-entering them.

  1. Given I am logged in with filters applied, when I tap "Save search", then the search appears in My Searches with its filters listed.
  2. Given I am logged out, when I tap "Save search", then I am sent to log in and returned with my filters intact.
  3. Given I already have 10 saved searches, when I try to save another, then I see "Limit reached" and nothing is saved.

Dependencies: none.

US-2: View and delete saved searches

Story: As a renter, I want to delete a saved search, so that my list only shows searches I still care about.

  1. Given I have saved searches, when I open My Searches, then I see each one with its name and date saved.
  2. Given I have no saved searches, when I open My Searches, then I see an empty state with a link to start searching.
  3. Given I tap delete and confirm, when the page reloads, then the search is gone.

Dependencies: US-1.

US-3: Get email alerts for new matches

Story: As a renter, I want an email when a new listing matches a saved search, so that I can contact landlords early.

  1. Given a listing is published that matches my search, when the daily job runs, then I get one email listing all new matches.
  2. Given no new matches, when the job runs, then no email is sent.
  3. Given the email bounces, when it fails 3 times, then alerts pause and I see a banner in the app.

Dependencies: US-1, email service setup.
Split suggestion: (a) matching job with logs only, (b) daily digest email, (c) bounce handling.

US-4: Unsubscribe in one click

Story: As a renter, I want to stop alerts from the email itself, so that I am not forced to log in.

  1. Given I click "Unsubscribe" in an alert, when the page loads, then alerts for that search stop and I see a confirmation.
  2. Given the link is older than 90 days, when I click it, then I am asked to log in.

Dependencies: US-3.

Open questions

  1. Daily digest only, or also instant alerts?
  2. Is the 10-search limit right for paid users?
  3. Does legal need a stated unsubscribe deadline?

Assumptions: email provider is already contracted; alerts are daily; size estimates assume a team of 4 engineers.

Written by AI from this prompt. Your results will vary.

Tips and common edits

Why this works

  • The one-user, one-outcome rule plus splitting on "and" stops vague stories that bundle several features together.
  • Requiring an edge or error criterion per story catches empty states and failures QA would otherwise find late.
  • Sizing with a Too large flag and split suggestion makes the output usable directly in sprint planning.

Use before backlog refinement or sprint planning. Paste your feature idea, rough notes or a Slack thread into ChatGPT, Claude or Gemini; know your users and sprint length so you can answer its questions.

Common edits

  • Swap S/M/L/XL for story points on your team's Fibonacci scale.
  • Ask for the stories as a CSV with columns ready to import into Jira or Linear.

Prompt details

By
Prompts-GPT editors
Published

Did this prompt work?

Keep exploring

More in Product Management