Prompting for Product Managers: delegate, describe, verify
Make AI a reliable tool for your PM deliverables: what to delegate, how to describe, how to verify, then an assistant set up for recurring tasks.
Course designed by Alex Dionisio, Senior Product Manager. Updated September 28, 2026.
- 54 steps
- ~2 hr 45 min
Level: Junior · Some basics are useful: introductory knowledge of the topic.
Turn generative AI into a reliable work tool for your Product Manager tasks: decide what to hand over, write prompts that give usable results on the first or second try, check what the AI produces, then set up a personal assistant for your recurring tasks. You leave with prompts tested on your own deliverables (feedback synthesis, persona, interview guide, user stories) and a configured assistant.
What you will be able to do
- Decide, for a task in your product cycle, whether to hand it to AI, in which mode (automation, augmentation, agent), and what you keep.
- Explain how an LLM behaves (prediction, dated knowledge, working memory, instruction following) to anticipate its mistakes and pick the right fix.
- Write a prompt that describes the deliverable, the process and the quality bar, with context, delimited data and examples when needed.
- Evaluate an AI output (content, process, behavior) against explicit criteria, then improve it through targeted iterations.
- Produce PM deliverables with AI (feedback synthesis, persona, interview guide, user stories) from real data, and check that they are faithful to it.
- Configure standing instructions and a dedicated assistant for a recurring task, test it on real cases, and decide when to move to an agent.
- Apply diligence rules (shared data, transparency, accountability) before using or circulating work produced with AI.
Prerequisites
- An account on a conversational assistant (Claude, ChatGPT, Gemini or similar); a free plan is enough for most exercises.
- Day-to-day product work (discovery, backlog, specs) or running a business project.
- No technical skills needed. To go further afterwards, "Build an AI assistant for your product" covers assistants built into a product, for your users.
Syllabus
What will I be able to do by the end of this course, and in what order?
- Course overview3 steps
Objective · By the end of this overview, you will know who the course is for, what you will produce (prompts tested on your own deliverables and a configured assistant) and how the three modules fit together.
How do you get a reliable result from generative AI, and know when not to trust it?
Objective · By the end of this lesson, you will be able to classify a task from your product cycle by the right delegation mode (automation, augmentation, agent, or no AI) and justify what stays under your responsibility.
Objective · By the end of this lesson, you will be able to trace an AI mistake back to its cause (prediction, knowledge, working memory, instruction following, sycophancy) and choose the matching fix instead of resending the same request.
Objective · By the end of this lesson, you will be able to write a prompt that describes the expected deliverable, process and quality bar, in separate blocks (role, context, task, data, format, criteria), and explain what each block contributes.
Objective · By the end of this lesson, you will be able to add to a prompt the technique that fits the problem at hand (examples, data delimiters, breaking the task into steps, permission to be uncertain) and justify your choice.
Objective · By the end of this lesson, you will be able to evaluate an AI output on three levels (content, process, behavior) against explicit criteria, and fix the prompt with a targeted change rather than a random rewording.
How do you apply these methods to discovery and delivery without losing fidelity to the data?
Objective · By the end of this lesson, you will be able to have a body of user feedback grouped into themes and opportunities, with each theme tied to verbatims you have checked, and explain what the synthesis does not tell you.
Objective · By the end of this lesson, you will be able to have a persona structured from real verbatims, separating what is observed from what is assumed, and list the hypotheses to validate in interviews.
Objective · By the end of this lesson, you will be able to produce a discovery interview guide from your hypotheses, then have it critiqued to remove leading, closed or hypothetical questions.
Objective · By the end of this lesson, you will be able to have user stories and acceptance criteria drafted from a validated opportunity, then have them critiqued (INVEST criteria, edge cases, open questions) before sharing them with the team.
How do you stop starting from scratch, hand more work to AI and stay accountable for the result?
Objective · By the end of this lesson, you will be able to write standing instructions suited to your role and choose between global instructions, a project and a conversation depending on how widely the context needs to apply.
Objective · By the end of this lesson, you will be able to configure an assistant dedicated to a recurring task, test it on real cases, and decide whether a task justifies an agent, with which human approval points.
- Working with diligence4 steps
Objective · By the end of this lesson, you will be able to decide which data to share with an AI tool, adjust its privacy settings, and apply a diligence checklist (data, checking, transparency, accountability) before circulating a deliverable produced with AI.
Glossary terms
Public definitions of the key concepts in this course:
Guide on this topic
- Evaluating an AI feature (evals)
Test cases, criteria, LLM-as-a-judge, release thresholds and monitoring, with an evals plan.
Verified sources
The fast-moving facts of this course (tools, models, regulation) are checked against these sources, then reviewed according to how quickly they go stale.
- Anthropic Privacy Center, How long do you store my data? · verified September 28, 2026
- Claude Academy, AI Fluency: Framework & Foundations · verified September 28, 2026
- Claude Academy, Introduction to Claude Cowork · verified September 28, 2026
- Claude Help Center, Get started with Claude Cowork · verified September 28, 2026
- Claude Help Center, Understanding Claude's personalization features · verified September 28, 2026
- Claude Help Center, What are projects? · verified September 28, 2026
- CNIL, Q&A on using a generative AI system (in French) · verified September 28, 2026
- Dakan and Feller, Framework for AI Fluency (Ringling College, CC BY-NC-ND 4.0) · verified September 28, 2026
- Google, Get started with Gems in Gemini Apps · verified September 28, 2026
- OpenAI Help Center, ChatGPT Custom Instructions · verified September 28, 2026
- OpenAI Help Center, ChatGPT Work and Codex · verified September 28, 2026
- OpenAI Help Center, Creating and editing GPTs · verified September 28, 2026