03 — Resources

AI Prompts for PMs

A practical prompt reference for product managers, operations leaders, and business users. The framework here is the one Elizabeth teaches in Unleashing Potential — built around the idea that the prompt is the key, and that better-written prompts open more doors.

"Think of your prompt as a key. The better it fits, the more doors it will open."

A — The nine steps

A framework for prompts that work the first time.

  1. 1. Be specific and detailed.

    Clearly define what you want. Instead of "tell me about cats," say "describe the behavior and characteristics of Siamese cats."

  2. 2. Start broad, then narrow down.

    Begin with a general context, then specify the exact question. Example: "Imagine you're teaching a 5th-grade science class. Explain the process of photosynthesis in simple terms."

  3. 3. Set the desired format.

    If you want bullet points, an essay, a dialogue, or a matrix — say so. Example: "List in bullet points the main causes of World War I."

  4. 4. Ask open-ended questions.

    Use "how," "why," or "what" instead of yes/no questions to get more comprehensive answers.

  5. 5. Consider multiple prompts.

    If you're not sure how to frame the question, try a few different phrasings. The best result usually comes from the third or fourth attempt.

  6. 6. Use iterative refinement.

    If the first answer isn't right, refine your prompt and ask again. AI conversations are a dialogue, not a search query.

  7. 7. Set the tone or style.

    "Explain quantum physics in a casual and humorous way." "Make this more formal and empathetic." Tone is part of the prompt.

  8. 8. Avoid being vague or ambiguous.

    Ambiguity gets you ambiguous answers. Don't ask about "climate change impacts" — ask about "the economic impacts of climate change on coastal real estate."

  9. 9. Test and learn.

    The more you use AI, the better you'll get at prompts. Save the ones that work. Adjust the ones that don't.

B — A worked example

From "tell me about CFO transitions" to a complete plan.

This is how the nine steps stack when you actually use them — each iteration adds one more dimension of specificity.

Pass 1 — Detail the task
"Detail the steps involved in a CFO transition."
Pass 2 — Add context
"In a corporate environment where a CFO is about to be replaced, outline a systematic approach to ensure a smooth transition."
Pass 3 — Set format & date
"…outline a systematic approach to ensure a smooth transition in a timeline format beginning July 10."
Pass 4 — Specify focus areas
"…emphasizing crucial milestones, activities, and aspects of team communication."
Final prompt
"In a professional manner, create a three-month timeline for a CFO transition in a corporate setting, emphasizing crucial milestones, activities, aspects of team communication, and engagement of department heads for the incoming CFO. Put into a matrix that includes titles of each section to ensure the organization of events with dates beginning July 10. Include meet-and-greet in Richland July 19, Seattle July 24, Portland July 26, and Boise August 2."

The output: a detailed transition timeline that saved hours of drafting. From here, you add specific people, additional details that come up, and any extra resources.

C — Multi-stage prompts

When the task is too complex for one prompt.

Break the prompt into stages, the same way you'd break a project into stages. Tell the AI to ask you for information in steps, then synthesize.

Example — Cover letter assistant
"Act as a career counselor. I'm about to apply for a job and I need to write a cover letter. You are going to ask me for information in 3 stages and then you will draft a cover letter. First ask me for my resume. Then ask me for the job description. Then ask me 5 questions related to the job description. Then compile all this information and draft a cover letter."
D — Everyday prompt tricks

Small phrases that change everything.

These are the prompts that turn AI from a novelty into a daily assistant.

E — Prompts that pay rent

Use cases worth keeping a library of.

Document work

"Please summarize the attached document into a one-page brief with three key takeaways and any decisions required from the reader."

Stakeholder messaging

"Write an email to [name] letting them know [situation]. Tone: professional, brief, no over-explanation. Sign off with [name]."

Market & competitor research

"Analyze the [industry] industry's key trends, competitors, and challenges in [year]. Provide industry insights, customer pain points, and opportunities."

Customer personas

"Create a B2B customer persona for [industry], including job roles, pain points, and content preferences. Make distinct personas for [segment A] and [segment B]."

Content planning

"Generate 10 LinkedIn post ideas for a [role] targeting [audience]. Include cost savings, industry trends, and thought leadership angles."

SEO content

"Rewrite this webpage content with SEO best practices while keeping it engaging. Use headings, structured data hints, and concise paragraphs. [paste content]"

Self-generating prompts (meta)

"Act as a prompt generator. I'll give you a title like 'Act as a code reviewer.' Give me back a prompt that's self-explanatory and appropriate to that title. My first title is: 'Act as a strategic finance advisor.'"
F — Cautions

It's a tool, not magic.

← Back to Resources