Still worried you are falling behind on AI as a PM? Don't worry. I've got you.
Last time I handed you a catalogue of Claude skills. A lot of you wrote back saying the same thing: the skills are great, but most of my AI work still happens in a plain chat box, and I never know what to type. Fair. Skills are the saved, reusable version. Prompts are the thing you reach for at 9 am when you just need the blank page to go away.
So I did the boring part, again (my local coffee place was really happy). I gathered 55 product management prompts and 7 full prompt libraries, sorted them by the work you actually do, and pulled out the five I would paste into a chat today without thinking twice.
One honest note before we go further, same as last time. This is a PM oracle. Some of these prompts are sharp. Some are basic. A few are barely more than a sentence. The value is not in owning all 55. It is in finding the three or four that quietly save you an hour a week and ignoring the rest.
Why prompts still matter in 2026
You would be forgiven for thinking prompts are old news. We have skills now. We have agents. We have models that hold a whole codebase in their heads. Why are we still talking about typing instructions into a box?
Because most of the work still starts in that box.
Skills are wonderful when you have a task you repeat the exact same way every week, and you are willing to set it up once. But a huge chunk of PM work is not like that. It is messy, one-off, and specific to the thing in front of you right now. A weird stakeholder situation. A pile of notes from a call that just ended. A PRD you need to pressure-test before a review in an hour. For all of that, a good prompt is still the fastest tool you own.
Here is the part people miss. A prompt is not magic words. It is a thinking structure. The reason a great prompt works is that it forces the right shape onto a vague request. "Help me prioritize" gets you mush. "Use RICE, ask for missing inputs, show your reasoning per feature, and give me a cut line" gets you something you can defend in a meeting. The prompt is doing what a good senior colleague would do: refusing to start until the question is framed properly.
My personal experience tells me that prompting is a little bit like working with an intern who doesn't know any better, but knows quite a lot. So, the more precise you are, the better examples you give, and the more time you spend actually asking what you need, the more likely you are to actually get what you want.
This is why prompt engineering works, not because those are magic words put into a secret Illuminati order. It's because someone did the work and figured out what makes a complete prompt for a given topic.
That is also why prompts have not been replaced by better models. A smarter model with a lazy prompt still gives you a lazy answer. The model got better at following instructions. It did not get better at reading your mind.
What does this actually change in your week?
Think about the work that fills your calendar but never makes it into your brag doc. Turning interview notes into themes. Drafting the same kind of PRD for the fifth time this quarter. Writing the exec update that says the same thing three different ways for three different audiences. Justifying a prioritization call you already made in your gut.
None of that needs genius. All of it takes hours.
That is the layer prompts attack. Not the strategy. Not the judgment. The repetitive scaffolding around the judgment. A good prompt library means you stop rebuilding that scaffolding from scratch every time. You paste, you fill in your specifics, you get a solid first draft, and you spend your real energy on the part that needs a human.
I want to be honest about the ceiling here, because the people selling you "10x your output" will not be. These prompts will not make you a better product thinker. They will give you back the time you currently spend on formatting, structuring, and staring at blank pages, so you have more room to be a better product thinker. That is a smaller promise. It also happens to be true.
The honest caveat: these are starting points, not answers
Every prompt below has blanks in it. The blanks are the whole point.
A prompt with {segment} and {problem_area} is not a finished tool. It is a frame waiting for your context. The output is only as good as what you put in, and what you put in is the part only you have: your users, your constraints, your data, your politics. Paste a prompt with lazy inputs, and you will get a confident, well-structured, useless answer. That is the worst kind, because it looks done.
So treat these the way you would treat a sharp intern's first draft. Read it with suspicion. Push back on it. The prompts that ask you clarifying questions first are the good ones, because they make you do the thinking that the output depends on. And as always, what works for me at my company may flop at yours. The only way to find out is to run one on real work this week and keep what earns its place.
The 5 prompts I'd paste today
I picked these because they cover the work nearly every PM repeats, regardless of seniority, company, or domain. For each one, I am giving you the raw template, a filled-in example so you can see it in action, a sample of what comes back, and why it earns a spot.
1. Turn messy notes into a real PRD
This is the blank-page killer. You have scattered thoughts from three conversations and a vague feeling of what needs building. This turns that into something you can hand to engineering. The reason it works is the first instruction: it asks before it writes.
The template
You are a principal product manager. Convert my messy notes into a clear
Product Requirements Document in markdown. First, ask up to 10 clarifying
questions. Then produce sections: Overview, Problem, Goals & non-goals,
User personas, Use cases, Requirements, UX notes, Open questions, and Risks.Filled in with real notes
[paste the template, then add:]
My notes: Users keep losing their search filters when they navigate away
and come back. Sales team complains the most. They want filters to persist.
Maybe make them shareable between teammates? Not sure about mobile. Eng is
worried about storage cost if we save this server-side for every user.What comes back (trimmed)
Before I draft, a few clarifying questions:
Should filters persist per device or across all of a user's devices?
Is "shareable" a must-have for v1, or a fast-follow?
What is the current drop-off you are seeing when filters reset?
Any data-retention or privacy constraints on storing filter state?
Is mobile in scope for this release or explicitly deferred?
Once you answer, here is the structure I will fill in:
Overview. Persisted search filters to reduce repeat work for high-frequency users, starting with the sales segment.
Problem. Filters reset on navigation, forcing users to rebuild context. Highest pain in the sales workflow.
Goals. Filters persist across sessions.
Non-goals. Cross-team sharing (deferred), mobile (deferred). ...
Why it matters: The clarifying questions are not a gimmick. They surface the decisions you were quietly avoiding, like whether mobile is in or out, before they become a fight in the review. You end up with a PRD and a sharper sense of your own feature.
2. Turn a pile of interview notes into themes
The most tedious job in the product. You did ten calls, you have pages of quotes, and somewhere in there are three insights that matter. Doing this by hand at 6 pm is how good research dies in a doc nobody reads. This is where AI genuinely beats the manual version.
The template
Act as a staff-level PM and qualitative researcher. I will paste raw notes
from multiple user interviews about '{topic}'. Cluster quotes into themes,
label each theme, count how many users mentioned it, and produce: (1) top 5
themes with intensity, (2) key quotes, (3) open questions for follow-up.Filled in
[template with {topic} = "first-week onboarding for our analytics tool",
then paste your raw interview notes below it]What comes back (trimmed)
Theme 1: Setup feels heavier than the payoff (7 of 9 users). Intensity: high. Representative quote: "I just wanted one chart, and I spent forty minutes connecting things."
Theme 2: Unclear what to do after connecting data (6 of 9). Intensity: high. Quote: "It connected, and then... now what?"
Theme 3: Templates would have saved me (4 of 9). Intensity: medium.
Open questions: Do the users who pushed through setup retain better than those who dropped? Is the friction the connection step itself or the empty state after it?
Why it matters: it does the clustering you dread and, more usefully, the counting. "Seven of nine mentioned this" is a sentence that wins roadmap arguments. The open questions at the end are the bit that keeps you honest, because the synthesis is a hypothesis, not a verdict.
3. Red-team your own PRD before someone else does
My favourite, and the direct cousin of the Idea Validator skill I shared last time. You wrote the PRD. You are too close to it. This prompt plays the skeptical senior person who finds the hole in the meeting, except it finds it for you first, in private, where fixing it is free.
The template
Act as a skeptical staff PM and tech lead reviewing this PRD. Identify
unclear assumptions, missing edge cases, risky dependencies, and misaligned
metrics. Return: (1) top risks, (2) questions for the author, (3) suggestions
to tighten scope while preserving impact.Filled in
[paste the template, then paste your PRD or a tight summary of it]What comes back (trimmed)
Top risks
Your success metric (activation rate) will move even if the feature fails, because of the seasonal signup spike. You cannot attribute cleanly.
The shareable-filters flow assumes a permissions model you have not specified. This is a hidden dependency on the access-control team.
No edge case for users on legacy plans who do not have saved-search at all.
Questions for the author
What happens to a shared filter when the owner's account is deactivated?
Is there a rollback plan if storage costs exceed the estimate?
Scope suggestions: Ship per-user persistence first, measure retention impact in isolation, and defer sharing until the permissions dependency is resolved.
Why it matters: It is cheap insurance. Three minutes of being told your metric is contaminated saves you a quarter of measuring the wrong thing. Use it on any PRD, strategy doc, or decision memo before it meets a real audience.
4. Prioritize with RICE, with the reasoning shown
Prioritization is a near-daily call, and the value of this prompt is not the score. It is that it forces you to defend the tradeoff out loud, with numbers, in a format you can put in front of a skeptical stakeholder.
The template
You are a product manager using RICE prioritization. I will provide a list
of features with notes. Ask for any missing inputs. Then create a table with
Reach, Impact, Confidence, Effort, and RICE score per feature, including your
reasoning in 1-2 sentences each. Finish with a ranked list and suggested cut line.Filled in
[template, then:]
Features:
1. Persisted search filters (sales team's top ask)
2. Dark mode (frequent but low-urgency request)
3. Bulk export (asked for by two enterprise accounts in renewal)What comes back (trimmed)
Feature | Reach | Impact | Confidence | Effort | RICE |
|---|---|---|---|---|---|
Bulk export | Med | High | High | Med | High |
Persisted filters | High | Med | High | Low | High |
Dark mode | High | Low | High | Low | Med |
Reasoning, per row, plus a ranked list:
Persisted filters (high reach, low effort, ties to a vocal segment)
Bulk export (lower reach but tied to renewal revenue)
Dark mode
Suggested cut line: below dark mode. Everything under it is a "later," not a "next."
Why it matters: the magic is "include your reasoning." A bare score is a black box you cannot defend. A score with a one-line rationale per feature is an argument. And when a stakeholder disagrees, you are now debating the inputs, which is a far better conversation than debating your vibes.
5. Draft the exec update you keep rewriting
Everyone writes these. The painful part is not the content, it is retuning the same progress for an exec who wants headlines, an engineer who wants detail, and a partner who wants reassurance. This handles the reframing.
The template
You are a PM writing an executive update. Using bullet points I provide,
craft a concise email with: context, 3-5 headlines, progress vs plan, key
risks, and asks. Keep it under 300 words, executive-friendly, and avoid
jargon. Offer two tone variants: neutral and slightly informal.Filled in
[template, then:]
Bullets: filters feature in dev, on track for end of month. Bulk export
slipped a week, waiting on security review. Need a decision on whether
mobile is in Q3 scope. Team morale good after last launch.What comes back (trimmed)
Subject: Product update: filters on track, export slipped a week
Quick update on where things stand.
Headlines: Filters on track for end of month. Bulk export slipped one week pending security review. One decision needed from you on Q3 mobile scope.
Progress vs plan: Two of three on schedule. Export delay is external (security queue), not engineering.
Risk: If the security review slips again, export moves to next month.
Ask: Can you confirm by Friday whether mobile is in Q3? It changes how we staff next sprint.
[Then a second, slightly warmer variant.]
Why it matters: the "two-tone variants" line is the quiet hero. You get to pick the register that fits your reader instead of agonizing over it. And the forced structure, headlines, then risks, then a specific ask, is exactly what a busy exec wants and what most updates bury.
Where to go deeper: 7 prompt libraries worth a look
If five is not enough and you want the full buffet, here are the libraries I pulled from. A mix of free and paid, single-model and tool-agnostic.
100+ AI Product Management Prompts, by Sid Saladi. Built for Claude and organized by 11 PM workflows. A strong free starting point. https://sidsaladi.substack.com/p/100-ai-product-management-prompts
PM Prompt Library, by Aakash Gupta. Paid, tool-agnostic, with a recommendation of which model to use per task.
https://growthpioneer.gumroad.com/l/zbensAI Prompt Library for Product Managers, by Product Leadership. A clean, model-neutral set aimed at product leaders.
https://www.productleadership.com/learning-hub/ai-prompt-library/AI PM Toolkit, by Erik Holmberg. An open-source GitHub repo of prompts and frameworks, handy if you live in code already.
https://github.com/erikholmberg/ai-pm-toolkit50+ ChatGPT Prompts for Product Managers, by CleverX. Broad coverage, written explicitly for ChatGPT.
https://cleverx.com/blog/chatgpt-prompts-for-product-managers/PM Prompt Library, by Kartik Sachdev. A meta-collection that points to many other good prompt sets.
https://survivalcrziest.github.io/ai/prompts/pm_prompt_library.htmlClaude PM Prompting Guide, a 40+ prompt guide built specifically for Claude, including discovery and assumption-testing techniques.
https://drive.google.com/file/d/1ucKkRf9IT4end6Fw_CpkTR0wdo8HP8VC/view
The full list
All 55 prompts, each with its category, author, target model, and the full copy-paste text, are in the sheet below. Sort it, search it, and steal the few that fit your week.
Category | Name | Full Prompt | Author |
Framework prompt (PRD) | Expert PM PRD Co-creator Prompt | You are an expert product manager. Work with me, the product owner, to generate a custom Product Requirements Document in markdown... | Switch Dimension |
Discovery – interview guide | Customer Interview Script Generator | You are a senior product manager. Design a 45-minute customer interview outline to explore the problem '{problem_area}' for the target segment '{segment}'. Include an intro script, 8–12 open questions, follow-up probes, and a short closing. Ask me 3 clarification questions before drafting. | Perplexity PM prompt set |
Discovery – synthesis | Interview Synthesis & Themes | Act as a staff-level PM and qualitative researcher. I will paste raw notes from multiple user interviews about '{topic}'. Cluster quotes into themes, label each theme, count how many users mentioned it, and produce: (1) top 5 themes with intensity, (2) key quotes, (3) open questions for follow-up. | Perplexity PM prompt set |
Discovery – JTBD | Jobs-to-Be-Done Extraction | You are a product strategist trained in Jobs-to-Be-Done. From the following interview notes about '{context}', extract 3–7 JTBD statements in the "When I… I want to… so I can…" format. Then list functional, emotional, and social jobs separately, and highlight underserved jobs with a short rationale. | Perplexity PM prompt set |
PRD – from notes | PRD from Messy Notes | You are a principal product manager. Convert my messy notes into a clear Product Requirements Document in markdown. First, ask up to 10 clarifying questions. Then produce sections: Overview, Problem, Goals & non-goals, User personas, Use cases, Requirements, UX notes, Open questions, and Risks. | Perplexity PM prompt set |
PRD – critique | PRD Red-Teaming Critique | Act as a skeptical staff PM and tech lead reviewing this PRD. Identify unclear assumptions, missing edge cases, risky dependencies, and misaligned metrics. Return: (1) top risks, (2) questions for the author, (3) suggestions to tighten scope while preserving impact. | Perplexity PM prompt set |
Roadmap & prioritization | Feature Prioritization Using RICE | You are a product manager using RICE prioritization. I will provide a list of features with notes. Ask for any missing inputs. Then create a table with Reach, Impact, Confidence, Effort, and RICE score per feature, including your reasoning in 1–2 sentences each. Finish with a ranked list and suggested cut line. | Perplexity PM prompt set |
Roadmap & prioritization | Outcome-Oriented Roadmap Draft | Act as a head of product. Using my goals, constraints, and backlog items, propose an outcome-oriented roadmap for the next {time_horizon}. Group work into themes, define quarterly outcomes, and map key bets to each outcome. Call out risks, dependencies, and what you would explicitly de-scope. | Perplexity PM prompt set |
User stories & tickets | User Story & Acceptance Criteria Generator | You are a PM writing implementation-ready tickets. From this feature description, generate 3–10 user stories in the "As a… I want… so that…" format plus detailed acceptance criteria and edge cases. Format output so it can be pasted into Jira, with one section per ticket. | Perplexity PM prompt set |
User stories & tickets | Bug Report to Repro Steps | Act as a QA engineer and PM pair. Given a messy bug report or Slack thread, extract the core issue, propose clear reproduction steps, expected vs actual behaviour, and impact assessment. Output in a template ready to paste into a bug tracker. | Perplexity PM prompt set |
Stakeholder communication | Executive Update Email Draft | You are a PM writing an executive update. Using bullet points I provide, craft a concise email with: context, 3–5 headlines, progress vs plan, key risks, and asks. Keep it under 300 words, executive-friendly, and avoid jargon. Offer two tone variants: neutral and slightly informal. | Perplexity PM prompt set |
Stakeholder communication | Slide Outline for Product Review | Act as a PM preparing a product review deck. Based on my notes, propose a slide-by-slide outline with titles and bullet points for each slide, covering: context, customer problem, solution overview, metrics, risks, and decision needed. Keep the outline to 8–12 slides. | Perplexity PM prompt set |
Metrics & analytics | North Star Metric Design | You are a product leader defining a North Star metric. From my product description, suggest 3–5 candidate North Star metrics and their supporting input metrics. For each, explain trade-offs, potential failure modes, and what behaviours it might unintentionally encourage. | Perplexity PM prompt set |
Metrics & analytics | Experiment Design & Hypotheses | Act as a data-informed PM. Given a product change idea, write 2–3 clear hypotheses, propose an experiment design (A/B, holdback, etc.), define primary and guardrail metrics, and list major risks to validity. Output as a one-pager draft. | Perplexity PM prompt set |
Strategy & vision | Product Strategy from Company Goals | You are a head of product. Using the company goals, constraints, and product context I provide, outline 3 strategic pillars, each with target customer segment, key jobs-to-be-done, and 2–3 example bets. Call out what we will deliberately not do in the next {time_horizon}. | Perplexity PM prompt set |
Strategy & vision | Competitor Teardown & Differentiation | Act as a competitive intelligence analyst and PM. From the product pages I paste, build a table comparing target users, key features, pricing, and positioning. Then suggest 3–5 differentiation angles and questions we should investigate before committing to a response. | Perplexity PM prompt set |
Execution – planning | Weekly PM Plan | You are my Chief of Staff. Based on my role, key projects, and calendar for next week, propose a focused weekly plan. Include 3 priorities, time-blocking suggestions, and what to explicitly drop or delegate. Keep the output to a one-page summary. | Perplexity PM prompt set |
Execution – retrospectives | Launch Postmortem Facilitator | Act as a facilitator for a product launch retro. From my notes and metrics, structure a postmortem into sections: What we aimed for, What happened, What worked, What failed, Surprises, and Concrete action items. Ask me for any missing context before finalizing. | Perplexity PM prompt set |
AI product work | AI Use-Case Finder for Existing Product | You are an AI-product PM. From my product description and constraints, propose 5–10 concrete AI use cases that improve user experience or efficiency. For each, describe user value, rough implementation idea (RAG, fine-tuning, agents, etc.), and main risks. | Perplexity PM prompt set |
AI product work | LLM Feature Risk Analysis | Act as a PM and AI safety reviewer. Given a proposed AI-powered feature, list possible failure modes (hallucinations, bias, abuse, privacy issues), their impact, and mitigations. Structure output as a risk register with severity and likelihood ratings. | Perplexity PM prompt set |
Stakeholder alignment | One-Pager for Exec Decision | You are a PM preparing a one-pager for an exec decision. Using my notes, draft a document with sections: Context, Problem, Options (with pros/cons), Recommendation, Impact, and Open risks. Keep it under 800 words and optimized for fast scanning. | Perplexity PM prompt set |
PRD – structure (external) | ChatGPT PRD Section Planner | ChatGPT, could you please suggest the key sections and relevant questions that should be included in a Product Requirements Document for a {type of product or feature}? | EarlyNode – ChatGPT prompts for senior PMs |
Competitive analysis (external) | Competitor Summary Prompt | ChatGPT, could you help summarise the key features, pricing, and user reviews of the following competitors: {list of competitors}? | EarlyNode – ChatGPT prompts for senior PMs |
Market trends (external) | Industry Trend Analysis Prompt | ChatGPT, based on your training data, what are the trending technologies, market shifts, or trends in the {specific industry}? | EarlyNode – ChatGPT prompts for senior PMs |
Customer feedback (external) | Customer Sentiment & Theme Finder | ChatGPT, could you please analyze these customer feedbacks {insert feedback} and highlight common patterns or recurring issues? | EarlyNode – ChatGPT prompts for senior PMs |
Roadmap planning (external) | Roadmap Planning Assistant | ChatGPT, based on these features {insert list of features}, can you help me create a product roadmap, suggesting potential milestones and considering dependencies between tasks? | EarlyNode – ChatGPT prompts for senior PMs |
Risk assessment (external) | Product Decision Risk Assessment | ChatGPT, considering these product decisions {list product decisions}, could you help identify potential risks, including technical debt, privacy issues, or market acceptance? | EarlyNode – ChatGPT prompts for senior PMs |
Stakeholder communication (external) | Stakeholder Update Drafting Prompt | ChatGPT, could you draft a communication to {insert stakeholder type} explaining {specific product decision or update}? Please make sure the language is tailored appropriately for the audience. | EarlyNode – ChatGPT prompts for senior PMs |
Market research (external) | Market Research & Validation | Conduct a market research analysis for [product idea]. Identify target users, demand, and potential competitors. | The AISurf – 10 Essential ChatGPT Prompts for PMs |
Roadmap planning (external) | 6-Month Roadmap Creator | Create a 6-month product roadmap for [product]. Include key milestones, feature releases, and prioritization strategies. | The AISurf – 10 Essential ChatGPT Prompts for PMs |
User personas (external) | User Persona Development Prompt | Develop detailed user personas for [product], including demographics, pain points, goals, and usage behavior. | The AISurf – 10 Essential ChatGPT Prompts for PMs |
Feature prioritization (external) | Framework Recommendation Prompt | Recommend the best framework (RICE, MoSCoW, Kano) to prioritize these features: [list]. | The AISurf – 10 Essential ChatGPT Prompts for PMs |
User stories (external) | User Story Writing Prompt | Write clear, actionable user stories for [feature]. Include acceptance criteria and edge cases. | The AISurf – 10 Essential ChatGPT Prompts for PMs |
GTM & launch (external) | GTM Launch Plan Prompt | Outline a go-to-market strategy for launching [product], including positioning, messaging, and acquisition channels. | The AISurf – 10 Essential ChatGPT Prompts for PMs |
Metrics (external) | Product Success Metrics Prompt | Define KPIs to measure the success of [product/feature] and explain alignment with business goals. | The AISurf – 10 Essential ChatGPT Prompts for PMs |
Feedback (external) | Feedback-Driven Iteration System | Design a system to collect and prioritize customer feedback to improve [product]. | The AISurf – 10 Essential ChatGPT Prompts for PMs |
Generic PM assistant | PM Brainstorming Partner #1 | You are my senior product management partner. Help me think through this decision: {decision_description}. Ask clarifying questions, surface trade-offs, propose 3–5 options, and suggest what additional data I should gather before committing. | Perplexity PM prompt set |
Learning & upskilling | Technical Concept Explainer #1 | You are a staff engineer teaching a product manager. Explain the concept of {technical_concept} in plain language, then give 3 examples in our product context and 5 questions I should ask engineers when this topic comes up. | Perplexity PM prompt set |
Discovery – surveys | Customer Survey Designer | You are a research-focused PM. Design a survey to evaluate the usability, performance, and perceived value of {product_name}. Include 10–15 questions mixing Likert scales, multiple choice, and a few open-ended questions, and group them into logical sections. | Perplexity PM prompt set (inspired by EarlyNode & GfG discovery prompts) |
Pricing & packaging | Pricing Plan Scenario Explorer | Act as a monetization PM. Given my current pricing model and 2–3 competitors, propose 3 alternative pricing or packaging scenarios. For each scenario, describe target customer segment, pros and cons, and key metrics we should monitor after launch. | Perplexity PM prompt set (inspired by GfG pricing prompts) |
Soft skills & leadership | PM Leadership Skill Plan | You are an executive coach for product leaders. Based on my self-assessment of strengths and weaknesses, propose a 90-day plan to strengthen 3 core PM skills (e.g., stakeholder management, storytelling, prioritization). Include weekly habits, resources, and reflection prompts. | Perplexity PM prompt set (inspired by Kraftful soft skills prompts) |
Framework application | AARRR Funnel Analysis for My Product | Act as a growth PM. Using the AARRR framework, map the customer journey for {product_description}. For each stage (Acquisition, Activation, Retention, Revenue, Referral), suggest key metrics, friction points to investigate, and 2–3 experiment ideas. | Perplexity PM prompt set (inspired by Kraftful AARRR prompt) |
Product documentation | API Integration Spec Helper | You are a technical PM writing an API integration spec. From my notes and an example request/response, draft a spec with: use cases, request/response schemas, authentication details, rate limits, error handling, and edge cases. Ask me for any missing technical constraints first. | Perplexity PM prompt set (inspired by public PM API prompts) |
Retrospectives & process | Sprint Retro Facilitator #1 | You are an agile coach facilitating a sprint retrospective for a cross-functional product team. Using my notes, organize discussion into: What went well, What didn't, Surprises, and Experiments for next sprint. Propose 3 concrete process improvements with owners and expected impact. | Perplexity PM prompt set |
Customer communication | Release Notes Draft Assistant #1 | You are a PM writing product release notes for customers. From my raw changelog, draft: (1) a concise customer-facing summary, (2) bullet list of key changes by user impact, and (3) one-sentence version for in-app announcement. Offer both technical and non-technical variants. | Perplexity PM prompt set |
Backlog grooming | Backlog Grooming Assistant | You are a PM running backlog grooming. From this unordered list of ideas and bugs, group items into themes, flag duplicates, and propose what to drop, delegate, or defer. Then suggest a trimmed list with clear next steps for each surviving item. | Perplexity PM prompt set |
Stakeholder alignment | Stakeholder Map & Influence Plan | Act as an organizational strategist. From my description of the project and people involved, map stakeholders by influence and interest, and propose a communication and alignment plan. Include messages, channels, and cadence for each key stakeholder group. | Perplexity PM prompt set |
Onboarding & UX | Onboarding Flow Heuristic Review | You are a UX-minded PM performing a heuristic review of our onboarding. Based on my screenshots or description, identify friction points, confusing copy, and missing moments of value. Return prioritized recommendations with expected impact on activation. | Perplexity PM prompt set |
Design collaboration | Design Collaboration Brief Helper | You are a PM preparing a brief for the design team. Using my notes, draft a design brief including problem, audience, constraints, success metrics, must-haves vs nice-to-haves, and research inputs. Keep it concise enough to paste into our design tool. | Perplexity PM prompt set |
Support & operations | Support Macro & Playbook Writer | Act as a PM partnering with support. From sample tickets and our product context, propose reusable response macros and a simple playbook for triaging similar issues in future. Include when to escalate to product or engineering. | Perplexity PM prompt set |
Change management | Change Announcement Drafter | You are a PM communicating a meaningful product change to existing users. Draft: (1) an internal FAQ, (2) an external announcement email, and (3) a short in-app banner copy. Address who is affected, what is changing, why now, and how to get help. | Perplexity PM prompt set |
Team effectiveness | Team Health Check Facilitator | Act as a facilitator for a team health check. From my notes about collaboration, pace, and morale, synthesize key themes and propose a workshop structure to address them. Include exercises, time boxes, and follow-up actions. | Perplexity PM prompt set |
Portfolio & multi-product | Product Portfolio Review | You are a CPO reviewing a product portfolio. Using my list of products, segments, and revenue contribution, identify overlaps, gaps, and cannibalization risks. Suggest 3–5 strategic moves, such as consolidations, sunsetting, or new bets. | Perplexity PM prompt set |
Data partnerships & integrations | Integration Opportunity Analyzer | You are a PM exploring integrations. From my user base, existing tools, and product strategy, list 10 potential integrations ranked by user value and feasibility. For top 3, outline a lightweight business case and key technical questions. | Perplexity PM prompt set |
Career & storytelling | PM Case Story Polisher | You are a PM career coach. From my rough notes about a project I led, craft a compelling STAR-style story (Situation, Task, Action, Result) that I can use in interviews or performance reviews. Propose 3 variations emphasizing different aspects (impact, leadership, ambiguity). | Perplexity PM prompt set |
You will not use most of them. That is fine. The goal was never to use 55 prompts. It is to find the handful that gives you an hour back, and to walk away from the rest without guilt.
Closing
AI is not going to replace product managers who use it well. But the PMs who quietly keep five good prompts within reach will spend less time fighting blank pages and more time on the work that actually moves the product.
Start with one. Paste it this week. See if it earns its place. Then decide what comes next.
If you try any of these, reply and tell me which one stuck. I read everything.
So, which blank page are you tired of staring at?
