ChatGPT Prompts for Product Managers: 12 PM Prompts

ChatGPT prompts for product managers work best when they turn notes you already have into stories, acceptance checks and a short roadmap slice. They should not invent metrics, quotes or launch dates. The 12 prompts below help you draft from facts, then hand the output to your team for confirmation.

The flagship is the product manager prompt for stories and roadmap from your notes. Paste the product name, audience, discovery notes and constraints, and it returns stories, Given/When/Then checks, a Now/Next/Later slice and a stakeholder update. For a tighter weekly stack of reusable PM prompts, see the prompt stack product managers actually reuse.

Capture discovery cleanly

1. Notes to problem bullets

Here are my raw discovery notes for [product]: [paste]. Turn them into 5 problem or opportunity bullets. Mark UNKNOWN where a claim has no support. Do not invent quotes.

2. Audience restatement

Restate who this change is for using only these words: [audience notes]. One sentence. No persona names I did not give you.

3. Constraint lock

List hard constraints from this planning text: [paste]. Separate must / should / out of scope. Do not add vendors or stack choices I did not name.

Write stories you can test

4. Story set from notes

From DiscoveryNotes: [paste], write 5 to 8 user stories in As a / I want / so that form. Use only roles present in the notes.

5. Acceptance criteria pack

For story [S#]: [text], write 3 to 5 Given / When / Then checks a QA engineer can run. No new analytics events unless I named them.

6. Edge cases for QA

Expand story [S#] into edge cases and empty states. Keep them testable. Link ideas to the test case generator from a user story if I need Gherkin next.

Plan the next slice

7. Now Next Later

Group these stories: [list] into Now / Next / Later for [horizon]. Do not assign points or calendar dates I did not supply.

8. Epic breakdown handoff

Turn this story list into an epic outline I can paste into Jira. Facts only. Keep story points and owners blank unless I supplied them. List the discovery questions I still need answered before the next epic.

9. Stakeholder update

Draft a short Slack update for [stakeholders] covering scope, Now items and open questions from [notes]. Label every UNKNOWN.

Keep the team honest

10. Metric hygiene

From OutcomesOrMetrics: [paste], list which numbers are real vs hoped-for. Refuse to invent a baseline.

11. Risk and dependency list

From KnownConstraints and notes: [paste], list risks and dependencies a PM should confirm this week. No drama language.

12. Pre-mortem questions

If this two-sprint slice failed, which gaps in my notes would explain it? Ask questions only; do not invent failure stories.

Worked example

A PM drafting TeamPulse pasted notes about Friday wins and blockers, a Monday digest request, a VPC storage rule and a one-screen design preference. The flagship prompt returned five stories, acceptance checks that reused only existing page_view analytics, a Now/Next/Later slice across two sprints and a stakeholder update that flagged the missing stale-blocker baseline. It refused NPS scores and competitor feature lists.

Variations

• Discovery only: run prompts 1 to 3 before any stories.
• QA heavy: pair prompt 5 with the test case generator page linked above.
• Launch messaging: after the slice is firm, move to the product marketing feature launch messaging brief.
• BA overlap: for as-is/to-be process maps, use the business analyst as-is to-be gap analysis.

Practical tips

• Paste notes, not vibes. Five concrete bullets beat a vague paragraph.
• Never let the model invent conversion rates or customer quotes.
• Send the stakeholder draft with a "needs confirmation" line on every UNKNOWN.
• Keep story points and dates in your tracker, not in the chat.
• TeamPulse is a made-up product. PromptDig has no ties to OpenAI, Atlassian, Slack or Notion.

Related prompts

Start from the prompt stack product managers actually reuse, then deepen with the test case generator from a user story, the business analyst as-is to-be gap analysis, AI prompts for business analysts, the product marketing feature launch messaging brief, ChatGPT prompts for product marketing and the landing page copy prompt with real proof.

Find more planning prompts in the Business prompts hub, browse every prompt on PromptDig, or share a prompt that helped you.