The PM Prompt Guide
Get more from every AI interaction
Most people use AI like a search engine. They type a question, get an answer, and either use it or complain it was not good enough. The people getting real leverage from AI treat it differently. This guide covers how.
The core principle
AI output quality is almost entirely determined by input quality. Bad prompts produce bad output. The model is not failing you — you are failing the model.
The most important shift is moving from questions to briefs. A question asks for an answer. A brief provides context, constraints, and a clear definition of what good looks like. The difference in output quality is not incremental. It is transformational.
The anatomy of a good prompt
Every effective prompt has four elements:
Role
Tell the model who it is. "You are a senior product strategist" produces different output than no role at all.
Context
What is the situation? What do you already know? What constraints exist? The more relevant context you provide, the less the model has to guess.
Task
Be specific about what you want. Not "write something about onboarding" but "write a 3-paragraph executive summary of the onboarding flow problems identified in these 5 customer interviews."
Output format
Tell it how you want the answer structured. Bullet points, a table, a narrative paragraph, a numbered list. If format matters to you, specify it.
Templates for PMs
Customer research synthesis
You are a senior UX researcher. I am going to share [N] customer interview transcripts. Identify the top 5 friction points mentioned, group them by theme, and note how frequently each appears. Output as a table with columns: Theme, Frequency, Representative Quote.
PRD first draft
You are a product manager writing a PRD for an engineering team. The feature is [description]. The problem it solves is [problem]. Key constraints are [constraints]. Write a PRD with sections: Overview, Problem Statement, Goals, Non-Goals, User Stories, and Open Questions.
Roadmap prioritization
I have the following [N] features to prioritize. For each, I will give you an estimated impact and effort score. Apply RICE scoring, explain your reasoning for each, and output a ranked list with a one-sentence rationale per item.
Executive update
You are a PM summarizing a sprint to a VP of Product. Keep it under 150 words. Cover: what shipped, what did not ship and why, top risk to address next sprint, and one customer signal worth noting. Tone: direct, no fluff.
Common mistakes
Being vague
"Make this better" tells the model nothing. Tell it what better means: more concise, more persuasive, better suited for a non-technical audience.
One-shot thinking
Treat AI like a conversation, not a vending machine. If the first output misses, refine it. "That is close but too formal — rewrite the second paragraph to sound more direct."
Accepting the first draft
AI first drafts are starting points. The model will hedge, over-qualify, and pad. Edit it down. Push it to be sharper.
No constraints
Unconstrained prompts produce generic output. Word limits, tone requirements, and format specifications are not restrictions — they are instructions.
Advanced moves
Once you have the basics, these techniques unlock another level of output quality.
Chain of thought
Ask the model to think step by step before giving an answer. "Before you respond, list your assumptions." This surfaces reasoning you can challenge.
Steelman opposing views
"What is the strongest argument against this approach?" Use AI to pressure-test your thinking before you present it.
Persona interviews
"You are a skeptical CFO reviewing this business case. Ask me the three hardest questions." Simulate stakeholder reactions before the real meeting.
Iterative refinement
"Here is the output. Here is what I do not like about it. Revise with these specific changes." Treat the session like a working relationship, not a transaction.
Need help putting this into practice?
Sprintt helps teams implement AI strategy, build product operating models, and ship faster. Reach out for a consultation.