A Week in My Life as a Product Leader with AI
I tracked every time I used AI in a work week - here are the 17 stand-out use cases
Dear subscribers,
Today, I want to share a week in my life as a product leader with AI.
After I tweeted the above, my DMs exploded with questions about specifics.
So here are 17 stand-out use cases after tracking every time I used AI at work:
This post is brought to you by…Lightricks
I recently created a short AI film inspired by one of my favorite games, Mass Effect.
This video took just 10 minutes to create using LTX Studio, which includes AI-assisted storyboards, consistent characters, epic music, and even SFX. I’m kind of blown away by how easy it is. Try creating your AI movie below.
Monday: Planning a new product
Summarize customer feedback. I start planning a new product by sharing rough notes with my customer community. I then paste the feedback (usually a messy chat thread) into AI to extract the takeaways.
Conduct market research. I love using Perplexity with DeepSeek R1 (free) and OpenAI Deep Research (costs $200 😱) for market research. I ask Deep Research for competitive analysis and tech stack trade-offs for my new product.
Draft a PRD. Instead of starting from scratch, I give AI my customer insights, market research, and initial thoughts. It creates a solid first draft based on my PRD template. Here’s a prompt if you want to try this yourself:
<customer> Paste customer feedback summary </customer>
<competition> Paste competitive research </competition>
<notes> Paste rough notes for the feature </notes>
<template> Paste your PRD template or a past PRD </notes>
Using the <customer> feedback, <competition> research, and my <notes>, help me draft an initial PRD based on my PRD <template>. Tuesday: Making design and technical trade-offs
Explore design variations. My designer shares a rough mock-up of the PRD that I wrote yesterday. I paste the mock into AI and ask it to prototype a few variations we then discuss with each other and with customers.
Review tech spec. My engineering lead drafts a tech spec for the same product. After reading it, I share the tech spec and my PRD with AI to get a second opinion on any gaps. I then discuss the results with my engineering lead.
Make technical trade-offs. Two engineers are debating a technical trade-off in a long Slack thread. Before jumping in, I share the full thread with AI to distill the core trade-offs. This helps me contribute more meaningfully to the discussion.
Tweak product copy. The designs are coming together, but something feels off about the product copy. I share the design with AI, along with my copy constraints. AI generates three copy variations that I then tweak manually. This is a super underrated use case (copy matters!), and here’s a prompt to get started:
Paste a screenshot of the design.
<goal> Describe what the design is trying to get the user to do </goal>
Please come up with three variations of the product copy in the screenshot that are more clear and concise given the <goal>.
Limit to (X) characters max.Course preview: The power of projects in giving AI the right context
Here’s another quick insight from my Become an AI-Powered Product Leader course:
I almost always talk to AI using a Project (Claude, ChatGPT) or Gem (Gemini). Dedicated workspaces with curated context dramatically improve the quality of AI responses.
For example, when planning a product, I create a Project loaded with my PRD, customer feedback, and meeting notes. This helps AI understand the full picture and give more nuanced responses.
In my course, I'll show you exactly how to structure these Projects for different scenarios, from career planning to interview prep.
Wednesday: Drafting team charter and strategy
Improve strategy doc. An adjacent team is revamping its strategy and wants my feedback. I think the doc focuses too much on tech instead of user problems. I ask AI to suggest edits to make it more focused on customer needs:
<doc> Paste the strategy doc</doc>
This <doc> is written too much from a tech perspective instead of focusing on user pain points.
Help me identify the gaps and suggest copy edits.Draft customer interview questions. The strategy could use more research on what customers want. I feed the document and my guide on how to talk to customers into AI and ask it to draft customer interview questions.
Brainstorm with AI voice. I'm still working on the strategy, but it’s time to go home. I load the draft document into AI and use its voice feature to brainstorm on the go.
Thursday: Building an AI product
Improve an AI prompt. I test the output of an early AI product with customers to identify issues. I share their feedback, the AI prompt, and my 7 advanced prompt techniques with AI to improve the prompt.
Edit evaluation criteria. A colleague has drafted evaluation criteria for the AI product. I ask AI to help make each criterion more precise and sort the list in priority order.
Run LLM evaluations. We use a state-of-the-art AI model to critique our AI product's output against our evaluation criteria. I audit the evaluation scores to maintain reliability.
For more on building AI products, check out my 3-part guide on picking the right AI use case, prompting/RAG/fine-tuning, and running evals.
Friday: Product updates, interviews, and performance reviews
Share product updates. I draft a product update on what my team is working on. Before sharing it in Slack, I feed it into AI to make it more concise based on my past product updates.
Improve self-review. It’s performance review season, and I share my self-review, my manager’s feedback, and the company performance matrix with AI. It helps me edit my word choices to be clearer and more aligned with expectations.
Edit peer feedback. I’m also behind on my 20 peer feedback requests, so I jot down some rough notes for a peer and asked AI to edit them for clarity and conciseness. I then manually use AI’s output to improve my peer feedback.
Synthesize interview notes. After interviewing a PM candidate, I share my raw interview notes and our hiring rubric with AI and ask it to clean up the notes. Here’s a quick prompt if you want to try something similar:
<notes> Paste interview notes </notes>
<rubric> Paste hiring rubric </rubric>
Help me edit these interview <notes> to align with our hiring <rubric>. Focus on concrete examples and clear signals. Keep it detailed but concise.Wrap up
The key is to treat AI as a co-pilot instead of an agent. I always manually refine AI’s suggestions rather than just paste them in.
In my upcoming AI course (March 29-30), I’ll dive deeper into how to 10x your productivity with my best AI workflows (see the outline below). Let me know what your favorite AI productivity hacks are in the replies!
Thanks for sharing! 🙏