ChatGPT prompts for product management

Act as a senior product manager helping me make a specific product decision. Do not assume that a loud customer request, a competitor feature, or an executive idea is automatically the right solution.

Context
Product and user: [PRODUCT AND USER]
Decision or opportunity: [DECISION]
Evidence available: [EVIDENCE]
Business and delivery constraints: [CONSTRAINTS]
Current goal and success metric: [GOAL AND METRIC]
Stakeholders and known disagreements: [STAKEHOLDERS]

Work in this order:
1. Restate the decision as a falsifiable problem statement. Separate the user problem, business problem, proposed solution, and assumptions; do not merge them.
2. Assess the evidence by source, sample, recency, and likely bias. Distinguish observed behavior from stated preference. Identify what would change the decision.
3. Define the target segment and job to be done. State who is deliberately out of scope.
4. Offer two or three credible options, including doing nothing or running a smaller test where appropriate. Compare expected user value, strategic fit, effort, dependencies, risks, and metric impact.
5. Recommend one option with a narrow first release. Write acceptance criteria, non-goals, instrumentation events, and a decision rule for expanding, revising, or stopping it.
6. List the next five actions, each with an owner role and the question it resolves.

Produce a 700-1,000 word product decision brief with sections: Decision, Evidence assessment, Options, Recommendation, First-release scope, Measurement plan, Risks, and Open questions. Use a compact comparison table for the options. Ask up to three clarifying questions only if a required input is missing; otherwise make conditional assumptions explicit.

Before answering, check that every recommendation traces to supplied evidence or a clearly labeled assumption. Also flag any metric that could improve while harming the intended user outcome.

Write a product requirements document for [FEATURE] in [PRODUCT] for [TARGET USER]. Base it only on [RESEARCH AND DECISION]. Include the user problem, job to be done, user flow, user stories, acceptance criteria, non-goals, edge cases, dependencies, analytics events, and launch criteria. Separate requirements from implementation suggestions. Produce 800 words maximum. Ask up to three questions only if an essential detail is absent. Check that each requirement supports the stated user problem and flag unsupported requirements.

Design a product experiment for [ASSUMPTION] affecting [USER SEGMENT] in [PRODUCT]. Available evidence is [EVIDENCE], and the constraint is [CONSTRAINT]. Specify the hypothesis, primary metric, guardrail metrics, population, comparison condition, minimum useful duration, qualitative follow-up, and decision thresholds. Include likely confounders such as seasonality, novelty, selection bias, and instrumentation gaps. Produce a one-page experiment plan. Do not invent baseline data; identify missing inputs and ask up to three questions.

Help prioritize [INITIATIVES] for [TIME PERIOD] in [PRODUCT]. Our strategy is [STRATEGY], capacity is [CAPACITY], and evidence is [EVIDENCE]. Score each item for customer problem severity, reach, confidence, strategic fit, effort, dependency risk, and reversibility, explaining each score. Do not use a single numeric score as the final decision. Produce a ranked table, a recommended portfolio, items to defer, and the evidence needed to revisit them. Flag initiatives that are solutions without a validated problem.