Insights · Issue 12 · From Meeta's desk

The PM's AI playbook: ten plays you can run this week.

A practitioner's manual for product managers using AI. Ten concrete plays, with copy-paste prompt skeletons, anti-patterns, and a verification step for each. Plus a week-in-the-life worked example and the stack to run it on.

Read the playbook ↓
Read time 14 min Author Meeta Vouk, Founder Topic PM craft · AI workflow · Prompts

If you read The PM Skill Map last month and walked away thinking "great map, now what do I actually do," this is the answer.

Ten plays. Each one is a workflow I or someone I trust runs every week. Each has a copy-paste prompt skeleton you can adapt to your context, a named anti-pattern, and a verification step. They are not everything a PM can do with AI — they are the highest-leverage ten. Get fluent in these and your hours-per-output ratio shifts in a way that compounds.

One principle first, because it determines whether any of these plays land.

The principle

Outsource effort. Keep judgment.

The PMs who get the most out of AI aren't the ones who use it most. They're the ones who know which parts of their job are effort — drafting, formatting, summarising, querying — and which parts are judgment — choosing, framing, deciding, owning. They outsource the first and protect the second.

Effort → give to AI

Drafting documents. Summarising calls. Writing SQL. Extracting themes. Generating variants. Categorising bug reports. Formatting outputs. Reading long things.

Judgment → keep yourself

Which problem to solve. Which framing to ship. Which tradeoff to make. Which signal matters. Which user to believe. Which risk to take. What "good" looks like.

Every play below is on the effort side of that line. The verification step is what keeps you on the judgment side. Skip the verification and the play stops working in about a week.

The ten plays

01

The Customer Feedback Synthesizer

When to use
After any research sprint — interviews, NPS responses, support tickets, sales-call transcripts, churn surveys. Anywhere you have 20+ pieces of customer voice and need to extract pattern.
Prompt skeleton
You're a senior product analyst. I'll paste [N] [interview transcripts / tickets / NPS comments] below.

For each one, extract:
- The job the customer was trying to do
- The pain or friction they described
- The workaround they're using today
- One direct quote worth saving

Then synthesize across all of them:
- Top 3 themes by frequency × severity
- Disconfirming evidence for our current roadmap
- Two surprising patterns I should investigate

Format: a table for the per-item extraction, then a one-page synthesis below.
[paste content]
Anti-pattern
"Summarize these interviews." — gives you mush with no decisions in it.
Verify
Spot-check 3 random items against the table. If extraction misses on more than 1, tighten the prompt before trusting the synthesis.
02

The PRD First Draft

When to use
You have a clear customer problem and a proposed solution. You need a working spec for engineering and design to push back on.
Prompt skeleton
You're an experienced PM at [company type]. I'll describe a problem and a proposed solution. Draft a PRD with:

1. Problem statement (1 paragraph)
2. Goals & non-goals (3 bullets each)
3. User stories (5-8, in INVEST format)
4. Acceptance criteria per story (Given/When/Then)
5. Edge cases (5+)
6. Open questions for eng + design
7. Success metrics (leading + lagging)

Be skeptical. Push back where the logic is thin.

Problem: [paste]
Proposed solution: [paste]
Anti-pattern
"Write me a PRD for X." — generic template back, useless.
Verify
Send the draft to one engineer and one designer. If they each return more than 5 unanswered questions, the prompt didn't carry enough context.
03

The User Story Writer

When to use
After a PRD or design review — you need INVEST-compliant stories for engineering, not the generic "as a user, I want…" stuff.
Prompt skeleton
Convert the following feature description into 5-10 user stories in INVEST format. Each story:

- Independent
- Negotiable
- Valuable (state the user value explicitly)
- Estimable (rough size tag: S / M / L)
- Small (fits one sprint)
- Testable (acceptance criteria in Given / When / Then)

Include 2 edge-case stories I might miss.

Feature: [paste]
Anti-pattern
Generic "as a user, I want…" with no acceptance criteria. Engineering can't build from those.
Verify
Read each story aloud. It should make sense as a Jira ticket by itself, with no surrounding context required.
04

The SQL Translator

When to use
You need data; you don't want to wait three days for an analyst. Most PM data questions are reachable with a 20-line query.
Prompt skeleton
I have a [Postgres / Snowflake / BigQuery] table named [name] with columns:
[list columns + types]

Write a SQL query to [specific question, with metric definition].

Make it readable, comment the joins, include a LIMIT for testing, and use CTEs where it helps clarity.
Anti-pattern
"Write SQL to get our funnel." — produces a query that runs but answers the wrong question.
Verify
Run on a sample. Compare one number to a known reference (yesterday's DAU, last week's revenue). Never ship analysis until at least one number matches what you'd expect.
05

The Funnel Diagnostic

When to use
A funnel step dropped. You don't yet know why. You want a structured hypothesis tree before you ask the data team.
Prompt skeleton
You're a senior PM. Here are funnel metrics for the last 4 weeks:

[paste data table]

Step [X] dropped from [Y%] to [Z%]. Generate:

1. Five hypotheses for why, ranked by likelihood
2. For each, the data I'd pull to confirm or rule out
3. For the top two, an experiment I could run in under a week
4. One non-data step worth doing in parallel (user interviews, etc.)
Anti-pattern
Asking the AI to "tell you why" — it doesn't know your business. It can hypothesize; you decide.
Verify
Take the top 3 hypotheses to your data team. If they laugh at any one, your prompt didn't include enough context.
06

The Competitive Scan

When to use
Quarterly competitive review. New vendor evaluation. Pre-board prep. Any moment you need to know what the competition is actually shipping.
Prompt skeleton
I need a competitive scan of [category] for [target buyer].

For each of these competitors:
[list 5-10]

Extract:
- Pricing model
- Their positioning statement, in their own words
- Strongest feature
- Weakest feature
- Any pricing or product change in the last 90 days
- Any hiring signal that suggests direction

Then synthesize:
- Where the market is moving
- Which two competitors are on the same trajectory
- Who is moving differently

Format: table + one-paragraph synthesis per competitor.
Anti-pattern
"Tell me about competitors" — Wikipedia-flavoured mush.
Verify
For two competitors, manually check the pricing claim against their website. AI hallucinates pricing constantly. If wrong, regenerate with explicit "use only what's in the linked content."
07

The Executive Briefing

When to use
You need to brief a CFO / CEO / Board in five minutes on a topic that took you five weeks to learn. The hardest writing in product.
Prompt skeleton
You're a strategic communications writer. I need to brief a [CFO / CEO / Board] in 5 minutes on [topic].

Output:
1. One-line top line (what they need to know)
2. Three supporting points, each one sentence
3. The one risk I'm worried about
4. The decision I need from them
5. The full one-page version, if they ask

Context: [paste your thinking]
Anti-pattern
A 12-bullet "summary." No exec reads 12 bullets.
Verify
Read the top line aloud. If it sounds like a headline a journalist would write, it's good. If it sounds like a sentence from a deck, rewrite.
08

The Bug Triage

When to use
A mountain of bug reports from support, beta, on-call — and the morning standup is in 20 minutes.
Prompt skeleton
Categorize the following bug reports. For each one extract:

(a) Component
(b) Severity (P0 / P1 / P2 / P3)
(c) One-line summary
(d) Is this a duplicate of an existing pattern?

Then group by component and severity. Flag the top 5 to look at first, ranked by: severity × user count × revenue impact.

[paste bug reports]
Anti-pattern
Reading the inbox sequentially, like email. Two hours later you've responded to ten, prioritized none.
Verify
For the top five, click into each one and confirm the severity matches your read. If the AI consistently over- or under-rates, recalibrate with a "P0 = revenue-blocking or data-loss" definition in the prompt.
09

The A/B Test Designer

When to use
You're about to ship a meaningful change and want to know if it actually worked. Designing the test poorly is the #1 reason teams ship false positives.
Prompt skeleton
I'm running an experiment:

Hypothesis: [paste]
Change: [paste]

Design it for me:
1. Primary metric (one — no more)
2. Guardrail metrics (3)
3. Sample size (baseline conversion [X]%, MDE [Y]%, alpha 0.05, power 0.8)
4. Allocation (50/50 or otherwise, with rationale)
5. Runtime estimate
6. Early-stopping criteria
7. Two plausible ways this could give a false-positive
8. What I'd do if it wins / loses / is inconclusive
Anti-pattern
Running the test, then deciding what to measure. Or measuring everything, which is the same thing.
Verify
Run the power calc with your actual data, not the AI's estimate. The AI's number is a sanity check; the real number is your job.
10

The Meeting-to-Action-Items Engine

When to use
After any 30+ minute meeting where you took notes (or had Granola / Fellow / Otter take them). Before the day ends.
Prompt skeleton
From these meeting notes, extract:

(a) Decisions made — with the one-line context for each
(b) Action items — with owner and due date
(c) Open questions — that didn't get resolved
(d) Watch-outs — anything I committed to that I might forget

Format:

DECISIONS
ACTIONS (owner · item · due)
OPEN (questions to revisit)
WATCH OUT (my commitments)

[paste notes]
Anti-pattern
"Summarize this meeting." — useless five minutes later.
Verify
Send the structured output (not the notes) to one other attendee. If they say "I don't remember agreeing to that" on any line, the AI hallucinated. Fix and re-send.

A week in the life

To make this concrete, here is one of my own weeks — lightly edited — with the plays mapped to the actual hours. The point isn't that you should copy the week. It's that once these plays are muscle memory, they slot into a normal PM week without adding hours. They subtract them.

A real PM week, with the plays slotted in

Hours saved are estimates against my pre-AI baseline two years ago. They underestimate; the compounding is real.

Monday~4 hrs saved
Weekend's customer interview backlog runs through Play 01 (Feedback Synthesizer) — 34 transcripts to a one-page synthesis in 25 minutes. Spotted a JTBD I'd been missing. Spent the saved time on a real conversation with the head of customer success.
Tuesday~3 hrs saved
New feature kickoff. Play 02 (PRD First Draft) takes the brief I wrote on the plane into a working PRD in 30 minutes. Engineering review the same day with real questions, not surface ones. Then Play 03 (User Story Writer) for the stories. Sprint planning runs on time for the first time this quarter.
Wednesday~2 hrs saved
Funnel metrics review. The signup-to-activation step is down 4 points. Play 05 (Funnel Diagnostic) generates 5 ranked hypotheses while I'm still in the dashboard. Play 04 (SQL) pulls the cohort-level data I need to confirm the top two. By lunch I have a working theory and a confidence interval.
Thursday~2 hrs saved
Quarterly competitive review with the GTM team. Play 06 (Competitive Scan) ran overnight on 8 named competitors. Spot-checked the pricing claims (one was wrong — regenerated). Output an updated battlecard before the meeting starts. Net: the meeting actually talked about strategy instead of who shipped what.
Friday~3 hrs saved
Board update due Monday. Play 07 (Executive Briefing) produces the one-pager from my running doc. Read the top line aloud — sounded like a journalist's headline, kept it. Play 10 (Meeting → Action Items) on this week's standups gave me my own to-do list for the weekend in two minutes.
All week~6 hrs saved
Play 08 (Bug Triage) runs every morning at 8am on the overnight inbox. Play 09 (A/B Test Designer) ran once on Wednesday's experiment proposal — caught a guardrail I would have missed.

Roughly twenty hours back, per week, doing the same job to a higher standard. The compounding shows up in the work that doesn't exist in the calendar — the strategy thinking, the customer calls, the time to actually read.

The stack — what to run all of this on

None of these plays require enterprise software. They require one or two well-chosen tools used well. Here is the minimum viable PM-AI stack as of mid-2026.

Layer Tool What it's actually for
LLM Claude or ChatGPT (pick one and go deep) All of Plays 01-03 and 05-10. Pay for the paid tier. The free model is for tourists.
Coding / data Cursor or Hex with text-to-SQL Play 04. Pasting SQL between a chatbot and your warehouse is for 2023. Live connection saves the verify step half its time.
Meeting capture Granola, Fellow, or Otter Feeds Play 10. Transcript-grade notes are now a commodity; pick one and use it consistently.
Customer voice Dovetail or raw transcripts → LLM Powers Play 01. If you have Dovetail, use it. If you don't, raw transcripts in a folder + a paid LLM gets you 80% of the way.
Competitive watch Crayon / Klue, or DIY with Browserbase + LLM Powers Play 06. The DIY version is what I run — half a day to set up, then it runs forever.
Experimentation Statsig, Eppo, or LaunchDarkly Plays in Play 09. The eval scaffolding here matters as much as the model.
Knowledge Notion or Confluence + LLM Q&A Where your PRDs, RFCs, retros, and customer-research artefacts live. AI gets useful in the company on day 90, when there's enough context for it to be wrong about.

Anti-patterns to avoid

What separates the PMs getting twenty hours back per week from the PMs getting frustrated is not the tools. It's whether they avoid the predictable ways AI workflows fail.

Anti-pattern 01
Using AI to draft what you can't yet think

If you can't sketch the answer on a napkin, AI can't help you. It is an amplifier of clarity, not a replacement for it. The output looks fluent and is therefore dangerous — you ship the model's version of "good enough" instead of yours.

Anti-pattern 02
Skipping the verify step

Every play has one. Every play. The plays stop working in about a week if you don't verify, because you build trust in outputs that haven't earned it. Verification is the practice that keeps you honest about what the model can and can't do this month.

Anti-pattern 03
Outsourcing judgment by accident

Asking the model "which feature should we prioritize" is not a play. It's an abdication. The model has no skin in your roadmap. Use it for the options; keep the choice.

Anti-pattern 04
One-shot prompting

The first response is rarely the best. Push back, ask for variants, ask for the opposite take. The PM who treats prompting as a conversation outperforms the PM who treats it as a vending machine, every time.

Anti-pattern 05
Pasting confidential data into a consumer tool

Customer names, revenue figures, internal documents — into the right model with the right enterprise terms, fine. Into a free consumer tier, not fine. Read your tool's data policy once. It will save you a meeting with legal later.

Anti-pattern 06
Hoarding the workflow

The plays only compound if your team learns them too. Pair on the prompts. Share the templates. The best PM AI workflows I've seen are team workflows — three or four people getting better at the same plays — not lone-wolf magic tricks.

The closing principle

The PMs I watch get the most out of AI share one habit. They treat their own prompts the way a good engineer treats their own functions — they iterate on them, version them, share them, and retire them when better ones come along. The prompt is the artefact. The output is just what the artefact produced this time.

So the homework, if you've made it this far, is small. Pick one of the ten plays. The one that maps to the thing on your calendar that hurts most this week. Run it. Verify it. Iterate the prompt twice. Save the version that worked into a personal "plays" doc.

By the end of the quarter you'll have ten of these. By the end of the year you'll have built a private operating system that compounds with you, week over week, that nobody can take away. That is the real return on AI for a PM in 2026. Not the demos. The plays.

M

Meeta Vouk

Founder, AI Impact Foundation. VP of Product at Teradata. Adjunct professor at NC State. 22 patents, 20+ years building enterprise AI — and a working belief that prompts are the new functions.

Practice these with us.

The AIIF Product Management track turns this playbook into hands-on practice — each play run on your own product context, with peer review and instructor feedback. You leave with a working personal operating system, not a certificate.

See the PM track →