Claude for Product Managers
Lesson 4
Writing Acceptance Criteria That Close the Interpretation Gap
The test for any acceptance criterion: Can a QA engineer verify this without asking you a question?
Prompt for writing better acceptance criteria:
Here are my user stories for [feature]:
[Stories]
For each story, write 3–5 acceptance criteria using this format:
GIVEN [context/state] WHEN [action] THEN [specific outcome]
Rules:
Every criterion must be objectively verifiable no subjective terms like "fast," "clear," or "appropriate"
Include at least one error state per story (what happens when something goes wrong)
Include at least one edge case per story (unusual but valid input)
Flag any story where you couldn't write a testable criterion that means the story itself needs to be more specific
Example output:
Story: As an operations manager, I want to search for tasks by name so that I can find items without knowing which project they're in.
Acceptance Criteria:
- GIVEN a user types at least 2 characters in the search bar, WHEN they press Enter or wait 300ms, THEN results appear within 1 second showing matching task names with the project name visible.
- GIVEN the search returns no results, WHEN the user views the results page, THEN they see the message "No tasks found for '[search term]'" and a link to create a new task.
- GIVEN a task name contains special characters (e.g., &, #, @), WHEN the user searches for it, THEN it returns correctly without error.
- GIVEN a user searches and a result task has been archived, WHEN results are displayed, THEN archived tasks appear with a visual indicator and open in read-only mode.