65 AI Prompts for Product Managers That Speed Up Every Sprint (2026)

The job description says strategy. The calendar says otherwise. Most PMs spend the week writing documentation, aligning stakeholders, synthesizing research they had no time to do properly, and explaining a roadmap that changed twice since the last all-hands.

These 65 AI prompts for product managers cover PRDs, user research, roadmaps, prioritization, GTM planning, sprint ceremonies, experiment design, competitive research, pricing, hiring, analytics, and more.

Built to handle the scaffolding so you can focus on the decisions.

How to Get Useful Output From AI as a PM

The problem with most AI outputs for PMs is not the model. It is that the prompt reads like a Jira ticket with no context. AI produces what you give it. If you give it a vague feature description, you get a vague PRD. If you want a deeper understanding of prompt types before diving in, the types of prompts for generative AI guide breaks down zero-shot, few-shot, chain-of-thought, and role-based prompts and when each works best. If you give it a detailed problem statement with user context and constraints, you get something worth building on.

What Makes a PM Prompt Actually Work

Three things separate a useful PM prompt from a waste of time. If you want broader context on how professionals across disciplines are using AI, the how professionals use AI chat guide is worth reading alongside this one.

Constraints. What are the non-negotiables? Platform limitations, engineering capacity, timeline, scope boundaries. Without these, AI makes optimistic assumptions that do not survive contact with your tech lead.

Audience. A PRD for an engineering team reads differently from a one-pager for the CEO. Specify who is reading it and why it matters to them.

Output format. Tell AI exactly what structure you need. Numbered sections, tables, bullet points, narrative prose. PMs who specify format get outputs they can drop directly into Notion or Confluence with minimal editing.

When Not to Use These Prompts

AI is not making your prioritization decisions. It is not telling you which problems are worth solving. It is not replacing user interviews or synthesis of qualitative signal. It handles the scaffolding. You bring the judgment.

AI Prompts for Writing PRDs and Feature Specs

The PRD is the document everyone reads differently and criticizes consistently. These prompts produce structured first drafts that give engineering and design something real to react to.

Full PRD and Feature Specification Prompts

Prompt 1: Full PRD

"Write a product requirements document for [feature name]. Include: problem statement (what user problem does this solve and what is the cost of not solving it), user personas affected, goals and non-goals, user stories in 'As a [user], I want [action], so that [outcome]' format (at least 5), acceptance criteria for each user story, edge cases and error states, success metrics with target values, dependencies, open questions, and out-of-scope items. Audience: engineering and design team. Format clearly with section headers."

What this is for: Any new feature or significant change. The non-goals and out-of-scope sections alone save more meeting time than anything else in this document.

Prompt 2: Lightweight feature brief

"Write a one-page feature brief for [feature idea] targeting [user persona]. Include: the problem in one sentence, the proposed solution in two sentences, what success looks like (2 to 3 measurable outcomes), the key constraints (technical, time, or resource), and the biggest open question that needs answering before we build. This is for an initial discussion, not final spec."

What this is for: Early-stage feature conversations where a full PRD is premature but you need something more than a Slack message.

Prompt 3: User story writing

"Convert this feature description into user stories with acceptance criteria: [paste feature description]. For each story write: the user story in standard format, 3 to 5 acceptance criteria as testable statements, the edge cases the story does not cover, and an estimated complexity (S/M/L). Flag any story that is too large to complete in a single sprint."

What this is for: Turning product decisions into engineering tickets that actually have enough detail to build from.

Prompt 4: Acceptance criteria from PRD

"I have this PRD section: [paste section]. Write detailed acceptance criteria for each requirement. Each criterion should be: specific enough that QA can write a test against it, written in plain language (no ambiguity), and cover both the happy path and at least one failure state. Flag any requirements that are too vague to write criteria for."

What this is for: PRDs that engineering will actually trust rather than ask twelve follow-up questions about.

Prompt 5: Edge case identification

"Review this feature spec and identify the edge cases and failure states we have not addressed: [paste spec]. For each one: describe the scenario, explain what the user would experience if we do not handle it, and suggest how we should handle it. Flag the top 3 that are most likely to occur and most likely to cause user frustration."

What this is for: The hour before engineering kick-off when someone asks "what happens if..." and you want to have already thought about it.

Prompt 6: API requirements document

"Write an API requirements document for [feature] that engineering needs to build. Include: the endpoints required, request and response formats, authentication requirements, rate limiting considerations, error codes and what they mean, and the data models involved. Audience: backend engineering team. Format as a technical specification."

What this is for: Features with significant backend work where engineering needs more than a feature description to estimate accurately. Use Chatly's AI Coder to generate boilerplate code snippets or API structure examples to include in the spec.

AI Prompts for User Research and Discovery

Discovery is where the best PMs invest the most time and where AI can process and synthesize faster than any human reading through transcripts.

Research Planning and Synthesis Prompts

Prompt 7: Interview guide

"Write a user interview guide for research on [problem area] targeting [user persona]. The research goal is: [state what you want to learn]. Include: 3 to 5 screener questions to qualify participants, a warm-up section to build rapport, 8 to 12 core questions using open-ended language (no leading questions), probing follow-ups for each core question, and a closing section. Flag any questions that risk anchoring the participant toward a predetermined answer."

What this is for: Structured discovery research where you need a guide that produces signal, not confirmation of what you already think.

Prompt 8: Research synthesis

"I have [X] user interview transcripts on the topic of [problem area]. Transcripts: [paste or describe]. Synthesize them into: the top 5 recurring pain points with direct evidence from transcripts, 3 unmet needs that appeared across multiple participants, 2 surprising findings that contradict our assumptions, and the jobs-to-be-done that users are trying to accomplish. For each insight, quote the specific transcript excerpt that supports it."

What this is for: Turning a folder full of transcripts into a structured synthesis document that the team can actually use to make decisions. Use Chatly's Chat PDF to upload transcripts directly and query them before running this synthesis prompt.

Prompt 9: Survey design

"Design a user survey for [research objective]. Include: 3 to 4 screener questions, 8 to 12 core questions using a mix of Likert scale, multiple choice, and open text, and a closing open-text question for additional feedback. Flag any questions that are double-barreled, leading, or ambiguous. Suggest the optimal order to minimize survey fatigue."

What this is for: Quantitative research at scale where you need signal across a large user base rather than depth from a small sample.

Prompt 10: JTBD mapping

"Map the jobs-to-be-done for [user persona] in [context]. For each job: write the functional job statement (what they are trying to accomplish), the emotional job (how they want to feel), the social job (how they want to be perceived), the current solution they use, the pain points with that solution, and the opportunity for improvement. Identify which job has the highest unmet need."

What this is for: Framing user problems in a way that opens up solution space rather than anchoring to the first solution that comes to mind.

Prompt 11: Usability test plan

"Write a usability test plan for [feature or product] with [user persona]. Include: research objectives (what specific questions we are testing), test tasks (5 to 7 scenario-based tasks written from the user's perspective, not feature names), what we will observe for each task, success criteria, and a discussion guide for the debrief. Format as a document a researcher or PM could run independently."

What this is for: Validating designs before engineering builds them. The task writing section alone is where most usability test plans fail.

Prompt 12: Customer feedback categorization

"I have [X] pieces of customer feedback from [source: support tickets, NPS responses, app reviews]. Raw feedback: [paste]. Categorize them into: product bugs, feature requests, UX friction points, and positive signals. For each category: list the specific feedback items, identify the most frequently mentioned theme, and score the urgency (High/Medium/Low) based on frequency and severity. Flag any feedback that contradicts our current roadmap assumptions."

What this is for: Making sense of unstructured feedback at scale rather than reading every ticket individually.

AI Prompts for Roadmap Planning and Prioritization

Prioritization is a decision. AI helps you structure the analysis. The call is still yours.

Prioritization Framework Prompts

Prompt 13: RICE scoring

"Score these [X] features using the RICE framework. For each feature: Reach (how many users affected per quarter), Impact (effect on key metric: Low=0.25, Medium=1, High=2, Massive=3), Confidence (percentage), and Effort (person-months). I will provide context for each. Calculate the RICE score for each, rank them, and flag any scores where my confidence estimate seems inconsistent with the evidence I have provided. Features and context: [paste]."

What this is for: Making prioritization discussions evidence-based rather than whoever argues loudest.

Prompt 14: MoSCoW prioritization

"Apply MoSCoW prioritization to this backlog for [product/team] for [time period]. Categorize each item as: Must Have (non-negotiable for launch/period), Should Have (high value, not critical), Could Have (nice to have, lower effort), Won't Have (explicitly not in scope). Constraints to apply: [engineering capacity, timeline, strategic goals]. Backlog items: [paste]. After categorizing, identify any conflicts where stakeholders are likely to disagree with the classification."

What this is for: Scope alignment conversations where you need a framework that forces explicit trade-off decisions.

Prompt 15: Opportunity scoring

"Score these opportunities using the opportunity scoring method. For each opportunity: importance to the user (1-10), current satisfaction with existing solutions (1-10), and calculate the opportunity score (importance + max(importance - satisfaction, 0)). Rank by score. Opportunities: [list]. Add a column for our current investment level in each area and flag misalignments between opportunity score and investment."

What this is for: Identifying where you are over-investing in already-solved problems and under-investing in high-opportunity areas.

Prompt 16: Roadmap narrative

"Write a roadmap narrative for [product] covering [time period: e.g. H2 2026]. The audience is [internal teams / board / customers]. Frame the roadmap around 3 themes: [theme 1], [theme 2], [theme 3]. For each theme: explain the strategic rationale, what we are building and why in that order, what we are not building and why, and how success will be measured. Do not make it a feature list. Make it a story about where we are going and why it matters."

What this is for: All-hands presentations, board updates, and customer-facing roadmap communications where a list of features does not land.

Prompt 17: Dependency mapping

"Map the dependencies for this set of planned features: [list features]. For each feature identify: which other features or infrastructure it depends on (blocking dependencies), which features depend on it (blocked dependencies), which external teams it requires (design, data, platform, marketing), and the minimum viable sequence to ship each feature. Flag any circular dependencies or scheduling conflicts."

What this is for: Roadmap planning conversations with engineering where sequencing matters as much as the feature list.

Prompt 18: Trade-off memo

"Write a trade-off memo on the following decision: [describe the decision]. We are choosing between [Option A] and [Option B]. For each option analyze: what we gain, what we give up, the risk profile, the reversibility of the decision, and the signal we will have in [X months] that tells us if we made the right call. End with a recommendation and the one assumption that would change it."

What this is for: Any significant product decision where you need to document your reasoning and align stakeholders before committing.

AI Prompts for Stakeholder Communication

The PM who communicates clearly to every audience is the one who gets the roadmap they want. These prompts help you tailor the message without starting from scratch each time.

Stakeholder Update and Alignment Prompts

Prompt 19: Executive product update

"Write a monthly product update for [executive audience: CEO, CPO, board]. Include: headline metrics vs targets with variance explanation, top 3 product wins this month, top 2 challenges and how we are addressing them, roadmap progress against plan, and the one decision we need from leadership this month. Under 400 words. Data-led, no spin."

What this is for: Keeping leadership informed without a 30-minute meeting every time the roadmap changes.

Prompt 20: Engineering alignment email

"Write an email to the engineering team explaining the rationale behind [product decision: e.g. a scope change, a pivot, a new priority]. Include: what changed and why, what this means for current work in progress, how we made this decision, what we are explicitly not changing, and where to ask questions. Transparent and specific. Under 300 words."

What this is for: Scope changes and priority shifts where the team deserves more than a Jira update.

Prompt 21: Stakeholder pushback response

"A stakeholder has pushed back on [product decision] with this argument: [paste their argument]. Write a response that: acknowledges their concern genuinely, explains the reasoning behind the decision using the data and context they may not have seen, addresses their specific objection directly, and proposes a way to incorporate their perspective without reversing the decision. Collaborative tone. Under 250 words."

What this is for: The inevitable pushback that comes when you say no to something a stakeholder wanted.

Prompt 22: Cross-functional kickoff brief

"Write a cross-functional kickoff brief for [feature/project]. Audience: engineering, design, data, and marketing. Include: the problem we are solving and why now, the proposed solution at a high level, each team's role and key responsibilities, the timeline and milestones, how decisions will be made when there is disagreement, and the definition of done. Under 500 words."

What this is for: Starting a cross-functional project with alignment instead of discovering misalignment three sprints in.

Prompt 23: Feature sunset communication

"Write internal and external communications for sunsetting [feature]. Internal version (for the team): why we are retiring it, what users will be told, the migration path if any, and the timeline. External version (for users): what is changing, why, what they should do next, and the deadline. Both versions should be honest without being defensive. Internal under 200 words, external under 150 words."

What this is for: Killing features cleanly without creating user backlash or internal confusion.

AI Prompts for Go-to-Market and Launch Planning

A product without a launch plan is a feature release. These prompts help you build the full GTM picture.

GTM Strategy and Launch Prompts

Prompt 24: GTM strategy brief

"Write a go-to-market strategy brief for [product or feature] launching in [timeframe]. Include: target customer segment with persona detail, the core value proposition in one sentence, primary and secondary channels with rationale for each, key messages for different audiences (users, buyers, press), launch milestones and timeline, success metrics and how we will measure them in the first 30 days, and the top 3 risks to the launch. Audience: cross-functional leadership."

What this is for: New product launches and major feature releases that need coordinated cross-functional execution.

Prompt 25: Launch checklist

"Create a launch checklist for [feature/product] going live on [date]. Include tasks across: product (feature flags, rollout plan, monitoring), engineering (load testing, on-call setup, rollback plan), design (final QA pass, empty states, error messages), marketing (copy live, channels ready, announcement scheduled), sales (trained, FAQs ready, demo updated), support (documentation live, escalation path clear), and data (tracking instrumented, dashboards ready). Add suggested owners and timing relative to launch day (T-14, T-7, T-3, T-0, T+7)."

What this is for: Every launch. The T+7 tasks are where most teams drop the ball.

Prompt 26: Positioning brief

"Write a positioning brief for [product/feature] targeting [buyer persona]. Include: the category we are competing in, the primary competitor or alternative we are displacing, the key differentiator in one sentence, the proof points that support the differentiator (3 to 5), the message that resonates with the [persona]'s primary pain point, and what we should never say or claim. This will be used by marketing and sales to build launch messaging."

What this is for: Ensuring marketing and sales are telling the same story with the same evidence. If you want prompts specifically for the sales side of a launch, the AI prompts for sales guide covers cold outreach, objection handling, and proposal writing in detail.

Prompt 27: Sales enablement FAQ

"Write a sales enablement FAQ for [feature/product]. Include 15 questions a prospect is likely to ask, organized into: product questions (what it does, how it works), competitive questions (how we compare to [competitor 1] and [competitor 2]), pricing and packaging questions, and implementation and support questions. For each question, write a concise answer under 75 words that a sales rep can use verbatim or adapt."

What this is for: Arming the sales team before a launch without a 2-hour training session they will forget by Friday.

Prompt 28: Beta program brief

"Write a beta program brief for [feature]. Include: the objective of the beta, the ideal beta user profile and how we will recruit them, what we are asking beta users to do and over what timeframe, how we will collect and act on their feedback, what success looks like for the beta, and the criteria for deciding to roll out broadly vs delay. Audience: internal team."

What this is for: Structured beta programs that produce actionable signal rather than random feedback from whoever opted in.

AI Prompts for Sprint Planning and Agile Workflows

Sprint ceremonies produce a lot of output and consume a lot of time. These prompts help you prepare for them properly.

Sprint Planning and Ceremony Prompts

Prompt 29: Sprint goal

"Write a sprint goal for a team working on [product area] for the sprint running [dates]. The work planned includes: [list key tickets or themes]. The sprint goal should: be achievable in one sprint, testable (we can confirm whether we hit it), meaningful to users or the business (not just 'complete X tickets'), and short enough to put on a sticky note. Write 3 options and explain the trade-offs between them."

What this is for: Sprint planning meetings where the goal ends up being "finish the backlog" unless someone pushes for something sharper. For the broader project management workflow beyond sprints, the AI prompts for project management guide covers risk registers, stakeholder reporting, and project closure documentation.

Prompt 30: Backlog refinement prep

"Help me prepare for a backlog refinement session. I have these items in the backlog: [paste tickets or descriptions]. For each item: assess whether it is ready to be refined (is the problem clear, is the solution defined enough, are dependencies identified?), list the questions that need answering before it can be estimated, and suggest whether it should be refined, broken down further, or returned to discovery. Flag any items that are too large for a single sprint."

What this is for: Coming to refinement with a pre-sorted backlog instead of spending the first 20 minutes figuring out what is on the list.

Prompt 31: Ticket writing

"Write engineering tickets for these features: [list features with brief descriptions]. For each ticket include: a clear title, the acceptance criteria as testable statements, the edge cases to handle, any relevant design or API links (placeholder), and the definition of done. Flag any ticket that needs to be broken down before it can be picked up in a sprint."

What this is for: The 30 minutes before sprint planning when you realize the tickets are not ready.

Prompt 32: Sprint review preparation

"Help me prepare for a sprint review for [team] covering the sprint ending [date]. Work completed: [list]. Outcomes vs sprint goal: [describe]. I need to present: a summary of what we shipped and why it matters, what we did not complete and why, any metrics that moved (or did not move), and what is coming next sprint. Write a structured sprint review script under 10 minutes when presented aloud. Audience: stakeholders who are not in the day-to-day."

What this is for: Sprint reviews that communicate outcomes rather than just listing what got done.

AI Prompts for Product Metrics and Experiment Design

The best PMs are not just data-informed. They design their work to generate useful signal. These prompts help with both.

Metrics Framework and A/B Test Prompts

Prompt 33: Metrics framework

"Define a metrics framework for [product area or feature]. Include: the North Star metric and why it is the right one, 3 to 5 primary metrics that drive the North Star, 3 to 5 secondary metrics for diagnosis, the counter-metrics that tell us when we are optimizing the wrong thing, how each metric is measured and where the data lives, and the reporting cadence for each level of the framework."

What this is for: Teams that are tracking everything and learning nothing because there is no hierarchy in their metrics.

Prompt 34: A/B experiment design

"Design an A/B experiment for [feature or change]. Include: the hypothesis in 'We believe that [change] will cause [outcome] because [rationale]' format, the primary metric, the minimum detectable effect we need to see, the estimated sample size required to reach statistical significance, the test duration based on current traffic, how we will handle novelty effects, the secondary metrics to monitor, and the stopping criteria. Flag any threats to validity."

What this is for: Experiments that produce actionable signal rather than inconclusive results you argue about for three weeks.

Prompt 35: North Star metric definition

"Help me define and pressure-test a North Star metric for [product]. The product does: [describe]. Our primary user is: [persona]. The core value the product delivers is: [describe]. Write 3 North Star metric candidates. For each: explain what it measures, how it connects user behavior to business outcomes, what it would miss or incentivize incorrectly, and how easy it is to measure with our current data infrastructure. Recommend one and explain why."

What this is for: The metric strategy conversation that every product team needs to have and most avoid.

Prompt 36: Funnel analysis brief

"Write a funnel analysis brief for [user flow: onboarding, activation, conversion]. Define each stage of the funnel, the metric that defines movement between stages, the current conversion rate at each stage (I will fill in), where the biggest drop-off is, 3 hypotheses for why users drop at that point, and the experiments we could run to test each hypothesis. Format as a brief I can share with data and engineering."

What this is for: Turning funnel data into an actionable investigation rather than a slide that says "conversion is low."

AI Prompts for Competitive Research and Market Analysis

Knowing the market is table stakes. Understanding where competitors are weak is the actual advantage.

Competitive Intelligence Prompts

Prompt 37: Competitor analysis

"Analyze [competitor name] as a competitor to [our product] in [market]. Include: their core positioning and target customer, their primary strengths, the areas where they are weakest or where users complain most (based on public reviews and signal I will provide), the features they have that we do not, the features we have that they do not, and the strategic gap they are leaving open that we could exploit. Sources I have: [list or paste]."

What this is for: Quarterly competitive reviews and pre-launch positioning work. Use Chatly's Ask AI to pull public information on competitors before running this prompt. Grok 4 is also strong here given its real-time awareness of market news and product announcements.

Prompt 38: Market sizing

"Help me build a market sizing estimate for [market or segment]. Use both a top-down and bottom-up approach. Top-down: start from [total addressable market], narrow to [serviceable addressable market] using [filter criteria], then to [serviceable obtainable market] using [our realistic reach]. Bottom-up: estimate [number of potential customers] multiplied by [average contract value or spend]. Show your assumptions explicitly and flag which ones are most uncertain."

What this is for: Investor conversations, strategic planning, and any decision that depends on understanding how big the opportunity actually is.

Prompt 39: Win/loss analysis

"Analyze these win/loss notes from recent sales and customer interviews: [paste]. Identify: the top 3 reasons we win deals, the top 3 reasons we lose deals, the competitor we lose to most often and why, the product gaps that appear most frequently in losses, and the customer profile of deals we win vs lose. Produce a structured summary the product and sales teams can act on."

What this is for: Closing the feedback loop between market signal and product decisions.

AI Prompts for Retrospectives and Post-Mortems

Retros produce action items nobody follows up on. These prompts are designed to produce fewer, better ones.

Retrospective and Post-Mortem Prompts

Prompt 40: Sprint retrospective facilitation

"Facilitate a sprint retrospective for a team that just completed a sprint with goal: [sprint goal]. Outcome: [met/partially met/not met]. Ask me 6 structured questions covering: what went well, what did not, the root cause of the biggest blocker, one process change to experiment with next sprint, and one team strength to double down on. After I answer, compile my responses into a structured retro report with a maximum of 3 action items."

What this is for: Retros that end with three concrete actions instead of a whiteboard covered in sticky notes.

Prompt 41: Feature post-mortem

"Write a post-mortem for [feature] that launched on [date] and [performed below expectations / had a significant incident / was rolled back]. Include: a timeline of what happened, the contributing factors (technical, process, decision-making), what we should have done differently at each stage, what we are going to change going forward, and the 2 to 3 most important lessons. Blameless and constructive. This will be shared with the broader product and engineering team."

What this is for: Learning from launches that did not go as planned without the conversation becoming defensive.

Prompt 42: Quarterly product review

"Write a quarterly product review for [team or product area] covering Q[X] 2026. Include: progress against OKRs with honest assessment of each (hit/miss/partial), the 3 most significant product decisions made this quarter and their outcomes, what we got wrong and what we would do differently, the biggest risks heading into next quarter, and the top 3 priorities for Q[X+1]. Audience: CPO and cross-functional leads. Under 600 words."

What this is for: Quarterly business reviews that actually drive strategic decisions rather than being a retrospective slideshow.

AI Prompts for Product Strategy and Vision

Strategy work is where PMs earn the most trust and spend the least time. These prompts help you build the thinking faster.

Strategy and Vision Prompts

Prompt 43: Product vision statement

"Help me develop a product vision statement for [product]. The product serves: [user]. It solves: [core problem]. In [X years], we want [product] to be: [ambitious but believable future state]. Write 3 vision statement options. Each should be: under 30 words, inspirational but grounded, and specific enough that the team can use it to make daily decisions. Avoid generic language like 'world-class' and 'best-in-class'."

What this is for: Strategy off-sites, new team onboarding, and any conversation where the team needs a shared picture of where they are going. Use Chatly's AI Story Generator to develop compelling product narratives that bring the vision to life for different audiences.

Prompt 44: Strategy one-pager

"Write a product strategy one-pager for [product] for [time period]. Include: the market opportunity in one paragraph, our strategic bets (3 maximum, each with a rationale), what we are explicitly not doing and why, the key risks and how we plan to mitigate them, and the metrics that will tell us if the strategy is working. Audience: executive team. Fits on one page."

What this is for: Getting strategic alignment without a 40-slide deck nobody reads past slide 12.

Prompt 45: Build vs buy vs partner analysis

"Analyze the build vs buy vs partner decision for [capability]. Consider: the strategic importance of this capability to our core product, the cost and time to build internally, the available vendor options and their trade-offs, the partnership opportunities and what we would give up, and the long-term implications of each path. Produce a structured recommendation with the key assumptions that would change it."

What this is for: Any significant capability decision where the answer is not obvious and the stakes are high.

AI Prompts for Customer Communication and Feedback Loops

PMs who talk to customers regularly make better decisions. These prompts make it easier to close the loop between what customers say and what gets built.

Customer Feedback and Communication Prompts

Prompt 46: Customer interview debrief

"Debrief this customer interview with me. I will describe what I heard. Ask me 5 questions to draw out the most important signal, identify any assumptions I might be making, and surface anything I might have missed. Then help me write a structured interview summary including: the customer profile, the key jobs they are trying to do, the pain points mentioned, any surprising moments, and the implications for our roadmap."

What this is for: Processing a customer interview while it is still fresh rather than letting the notes sit in a folder.

Prompt 47: Product changelog entry

"Write a changelog entry for [feature or update] released on [date]. Audience: existing users who want to know what changed and why it matters to them. Include: a clear headline of what changed, a one-paragraph explanation of why we built it (frame it from the user's perspective, not ours), how to access or use it, and any known limitations in this version. Under 150 words. No corporate announcements."

What this is for: Changelogs that users actually read because they explain the why, not just the what.

Prompt 48: NPS follow-up outreach

"Write an outreach message to a user who gave us an NPS score of [6 or below] with this comment: [paste comment]. The message should: acknowledge their specific feedback without being defensive, ask one follow-up question that would help us understand the problem better, and invite them to a brief conversation if they are willing. Under 100 words. Human and genuine, not a support ticket response."

What this is for: Closing the loop with detractors in a way that might actually surface useful signal. Run the message through the Chatly AI Humanizer before sending to make sure it reads like a person wrote it.

Prompt 49: Feature request response

"Write a response to a customer who has requested [feature]. We have [decided not to build it / put it on the roadmap / are investigating it]. The response should: acknowledge their request and the problem it represents, explain our current thinking honestly, give them something useful even if it is not the feature they asked for, and leave them feeling heard rather than brushed off. Under 150 words."

What this is for: Feature request management that does not feel like a form letter.

Prompt 50: User onboarding email sequence

"Write a 5-email onboarding sequence for new users of [product]. Each email has one job. Email 1 (day 0): confirm the account and deliver one immediate win. Email 2 (day 2): address the most common early obstacle. Email 3 (day 5): introduce the feature most correlated with long-term retention. Email 4 (day 10): social proof from a user with a similar profile. Email 5 (day 14): check-in and surface the next level of value. For each: subject line, preview text, and body under 150 words."

What this is for: Activation sequences that move users from signup to habitual use rather than churning in week two.

Prompt 51: Deprecation announcement

"Write a user-facing deprecation announcement for [feature being deprecated]. Timeline: [date feature goes offline]. Include: what is changing, why (honest but not technical), what users should do before the deadline, what will happen to their data, where to go for help, and a clear deadline. Under 200 words. Clear and direct without being alarming."

What this is for: Deprecation communications that users actually act on rather than ignore until the feature disappears.

Prompt 52: Roadmap update for customers

"Write a customer-facing roadmap update for [quarter]. We are sharing: [list what we shipped], [list what we are working on next], and [list what we decided not to build and why]. Frame this as a progress update that shows we are listening to customer feedback and making deliberate choices. Transparent about trade-offs. Under 300 words. This will be sent to our top 50 customers."

What this is for: Keeping high-value customers informed and engaged with your product direction without overpromising on timelines.

AI Prompts for Pricing and Packaging Strategy

Pricing is a product decision. Most PMs treat it as a finance decision. These prompts help you think through it properly.

Pricing Research and Strategy Prompts

Prompt 53: Pricing model analysis

"Analyze the pricing model options for [product] targeting [customer segment]. Evaluate: per-seat pricing, usage-based pricing, tiered feature packaging, freemium with upgrade triggers, and flat-rate subscription. For each model: explain how it aligns with how customers derive value, the revenue predictability it offers, the operational complexity it creates, and which customer behaviors it incentivizes or disincentivizes. Recommend a model and explain the key assumption that would change the recommendation."

What this is for: Pricing strategy decisions where the team is defaulting to "what competitors charge" instead of thinking from first principles. For the financial modeling side of pricing decisions, the AI prompts for financial analysis guide has prompts for scenario planning and revenue modeling.

Prompt 54: Packaging brief

"Write a packaging brief for [product] with [X tiers]. For each tier: the target customer segment, the core features included, the features deliberately excluded and why, the price point rationale, the upgrade trigger that should move a customer from this tier to the next, and what a customer in this tier looks like 12 months after purchase. Audience: product, marketing, and sales."

What this is for: Turning a pricing spreadsheet into a packaging strategy that sales can actually use to qualify and upsell. Use the Chatly Paraphrasing Tool to adapt the same packaging description for sales decks, landing pages, and internal briefs without rewriting each from scratch.

Prompt 55: Price increase communication

"Write the internal brief and customer communication for a price increase for [product]. Internal brief: the rationale, the expected churn impact, the segments most at risk, and the retention strategy for high-value customers. Customer communication: honest explanation of the change, what they are getting for the higher price, the timeline, and any grandfather provisions. Internal under 300 words. Customer-facing under 150 words."

What this is for: Price increases that do not become a churn event because the communication was an afterthought.

AI Prompts for PM Hiring and Team Building

If you are a lead PM or head of product, hiring is part of the job. These prompts make it faster without making it worse.

Hiring and Team Development Prompts

Prompt 56: PM job description

"Write a job description for a [seniority level] Product Manager role at [company type] in [industry]. Include: a 2-sentence role summary that explains the impact of the role (not the admin of it), 5 key responsibilities framed as outcomes not tasks, required qualifications that are honest not aspirational, 3 nice-to-haves, and a culture statement specific enough to put someone off who would not thrive here. Avoid: 'fast-paced environment', 'self-starter', 'data-driven', 'passionate about product'."

What this is for: Job descriptions that attract people who can actually do the job rather than people who know how to describe the job.

Prompt 57: PM interview scorecard

"Write a PM interview scorecard for a [seniority level] product manager role. Include: 5 core competencies to assess (e.g. problem framing, prioritization, stakeholder influence, data reasoning, communication), the interview question or exercise that tests each competency, and a 4-level rubric (Strong No, No, Yes, Strong Yes) with specific behavioral indicators at each level. Make the rubric specific enough that two interviewers with different styles would score the same candidate similarly."

What this is for: Structured hiring that produces consistent signal across a panel rather than vibes-based decisions.

Prompt 58: PM performance review

"Write a performance review for a [seniority level] PM who has demonstrated [key strengths] but needs to develop in [development areas]. Include: 3 specific examples of strong product judgment with business impact, 2 specific examples of where development is needed with the exact behavior to change, 2 measurable development goals for the next review period, and an overall assessment. Constructive, direct, growth-oriented. Under 500 words."

What this is for: Performance conversations that actually help a PM grow rather than checking an HR box.

AI Prompts for Product Documentation and Knowledge Management

Good product teams have a shared understanding of decisions and context. These prompts help you build and maintain it.

Documentation and Knowledge Prompts

Prompt 59: Decision log entry

"Write a decision log entry for [product decision made on date]. Include: the decision made, the options that were considered, the data and reasoning that drove the choice, who was involved in the decision, the assumptions being made, the expected outcome, and how we will know if it was the right call. This will be stored in our product wiki for future reference. Under 300 words."

What this is for: Building institutional memory so future PMs understand why decisions were made, not just what was decided.

Prompt 60: Product FAQ for new team members

"Write a product FAQ document for a new PM, designer, or engineer joining the [product area] team. Anticipate the 12 questions someone new would have in their first two weeks: about the product, the users, the current strategy, the tech stack (high level), the team norms, and the most important context from decisions made before they joined. Answers under 60 words each. Honest, not a sales pitch for the product."

What this is for: Onboarding new team members without a two-week shadowing process that pulls the whole team away from their work.

Prompt 61: Meeting notes to action items

"Convert these meeting notes into a structured output: [paste notes]. Produce: a 3-sentence summary of what was discussed, decisions made (numbered list), action items with owner and deadline for each, open questions that need resolution, and any dependencies or blockers surfaced. Flag any action items that are missing an owner or deadline."

What this is for: Every meeting that produces decisions and actions. The ten minutes you spend doing this saves hours of follow-up confusion.

AI Prompts for Product Analytics and Reporting

Data literacy is table stakes for PMs. These prompts help you move from raw numbers to actual insight faster.

Analytics and Reporting Prompts

Prompt 62: Weekly product metrics report

"Write a weekly product metrics report for [product area] covering the week of [date]. Metrics data: [paste]. For each key metric: current value, change vs prior week (absolute and percentage), whether it is trending in the right direction, and the most likely explanation for any significant movement. Highlight the one metric most worth discussing in the weekly product review and explain why. Under 400 words."

What this is for: Weekly reporting that surfaces signal rather than just listing numbers nobody acts on.

Prompt 63: Cohort analysis brief

"Write a cohort analysis brief for [user cohort: e.g. users who signed up in Q1 2026, enterprise users activated in January]. Include: how the cohort is defined, the key behaviors we are tracking over time, the retention or engagement curve description, how this cohort compares to previous cohorts, and what the data suggests about product changes that may have driven the difference. Data: [paste or describe]."

What this is for: Understanding whether product changes are actually improving user outcomes over time, not just in the first session.

Prompt 64: Data request for engineering or data team

"Write a data request for the [data/analytics/engineering] team. We need: [describe the data needed], to answer this question: [the product question]. The data should cover: [time period], [user segment], [relevant events or attributes]. Output format needed: [table, dashboard, CSV, API]. Priority: [High/Medium/Low] because [reason]. We need it by [date] because [reason]. Please flag if any of this data is not currently tracked."

What this is for: Data requests that actually get prioritized because they are specific, justified, and easy to act on.

Prompt 65: OKR check-in update

"Write a mid-quarter OKR check-in update for these objectives: [paste OKRs with current metric values]. For each Key Result: current status (On Track/At Risk/Off Track), the evidence supporting that status, what is driving the result (positive or negative), what we are doing about it if at risk, and any changes to the target we are proposing and why. Audience: CPO and cross-functional leads. Honest, not optimistic."

What this is for: OKR reviews that produce real accountability rather than green status updates on red situations.

ChatGPT vs Claude vs Gemini for Product Management

Different PM tasks pull different models. Here is where each one earns its keep.

Which Model Handles Structured Documents Better

GPT-5 is the most reliable model for structured PM documents. PRDs, launch checklists, sprint tickets, strategy one-pagers. It follows multi-part formatting instructions without dropping sections, handles complex nested structures, and produces output that lands in Confluence or Notion without major cleanup. For any document that needs to be comprehensive and well-organized, GPT-5 is the default starting point. For complex strategy documents and long-context analysis, Claude Opus 4.6 is worth testing alongside GPT-5.

Which Model Handles Narrative and Communication Better

Claude Sonnet 4.6 is stronger for anything that needs to persuade or be read by a non-technical audience. Executive updates, stakeholder alignment emails, customer communications, feature announcements. Claude writes with more natural flow and picks up on tone nuance that GPT-5 sometimes misses. For anything going to the CEO or a customer, start with Claude.

Which Model Is Useful for Research and Current Context

Gemini 2.5 Pro is useful when your PM work requires real-world context, market data, or current competitive information. For competitive analysis prompts where you want the model to bring recent context, or for market sizing where current data matters, Gemini's broader awareness is an advantage.

The Case for Running the Same Prompt Twice

The PRD that GPT-5 writes will be more complete. The executive one-pager that Claude writes will be more readable. You do not have to choose if you are using Chatly. Run the prompt on both, take the structure from one and the tone from the other, and spend the time you saved on actual product thinking.

Why Chatly Is the Practical Answer

Comparing models is only useful if comparing them is easy. Chatly puts GPT-5, Claude, Gemini, Grok, and 30+ other models in one interface. Write the prompt once. Run it across the models that suit your task. Use the output that works. That is the whole advantage, and it takes seconds rather than the minutes you would spend copying between separate tools.

How Chatly Fits Into a PM Workflow

PMs use more tools than almost any other role and spend more time context-switching between them. Chatly does not replace Jira, Notion, or Figma. It sits at the point where you need to produce a document, synthesize information, or communicate a decision and removes the blank page problem in every one of those moments.

Chatly gives you access to 30+ AI models including Kimi K2 and Deepseek R1 in one interface. Write a prompt once, run it across the models best suited for that task, compare outputs, use the best one. For PM-specific tasks, the Summary Generator is useful for condensing long research documents or stakeholder notes before briefing AI. The Paraphrasing Tool adapts the same core message for different audiences without rewriting from scratch. For teams working with PDF documents like research reports, analyst briefs, or competitor teardowns, Chat PDF lets you query them directly.

If you want to see how AI fits into broader professional workflows beyond product, the AI prompts for business guide covers the organizational layer that most PMs also touch. Chatly offers a free plan, with the latest pricing always on the Chatly.

Conclusion

The constraint on most PM work is not ideas or even prioritization frameworks. It is time. The PRD that takes three days to write delays engineering by three days. The stakeholder update that never gets sent creates a misalignment that costs a sprint to fix. The discovery synthesis sitting in a folder because nobody had time to write it up means the next quarter's roadmap is built on assumptions instead of signal.

These prompts do not make the decisions. They remove the document production bottleneck so the decisions can actually happen. Fill in the brackets with real context, run the prompt, edit the 20% that needs your judgment, and move on to the next thing.