AI prompts for PM work
Product-management threads shared prompt templates that simulate senior PM thinking — covering positioning, ICP, MVP features, user journeys, RICE prioritization and roadmaps formatted for design handoff. Multiple social posts and examples show prompts that turn docs into decision memos or generate full mobile app plans to speed side-project execution (x.com, x.com, x.com).
Product managers are swapping one-off chatbot questions for reusable prompt templates that draft roadmaps, score features, and turn rough notes into planning docs. (openai.com) The posts circulating this week show a familiar pattern: assign the model a senior product role, paste in product context, then ask for outputs like ideal customer profiles, minimum viable product feature sets, user journeys, RICE scoring, and design-ready handoff docs. OpenAI’s prompting guide says output quality improves when instructions are clear, grounded in reference material, and formatted for a specific response. (openai.com) Anthropic’s prompt-engineering guide makes the same case from the model side: better results come from explicit roles, examples, structured formatting, and prompt chaining, which is splitting one big task into smaller linked steps. Several PM prompt libraries now package that approach into copy-and-paste workflows. (anthropic.com) That maps neatly onto how product work is already organized. Intercom’s RICE framework scores ideas by reach, impact, confidence, and effort, and prompt libraries now ask models to fill in those fields, compare tradeoffs, and rank what to build first. (intercom.com) The newer shift is format, not just speed. PM Toolkit says its workflows chain prompts across launch planning, quarterly planning, feature discovery, pricing, and churn analysis, while Glean’s January 27, 2026 prompt list includes customer-pain summaries, competitor review comparisons, release plans, and roadmap presentation drafts. (pmtoolkit.ai) (glean.com) Those examples show why the posts are spreading beyond full-time product teams. A solo builder can paste a market idea into a template and get a positioning memo, a feature list, a release sequence, and a first-pass requirements document in minutes instead of building each artifact from scratch. (glean.com) (pmtoolkit.ai) The tradeoff is that polished output can hide weak assumptions. OpenAI tells users to provide grounding context and test prompts systematically, and Anthropic recommends examples and iterative refinement rather than treating the first answer as final. (openai.com) (anthropic.com) That leaves the real value in the template, not the flourish. The most useful PM prompts are acting less like magic commands and more like checklists that force founders and product managers to state the customer, the problem, the tradeoffs, and the next release in plain terms. (openai.com) (intercom.com)
