Twice a year I sit down with each person on my team and talk about their skills. For product managers, I use Petra Wille’s PM Wheel. I love it.

The PM Wheel is simple. It’s a circle with eight spokes, each one a skill a product manager needs. You score yourself on each spoke from 0 at the center to 8 at the edge, then connect the dots. What makes it work is the set of questions behind each skill. "Rate your stakeholder management" is a meaningless question; everyone gives themselves a six. Answer twenty specific questions about how you actually manage stakeholders and the number means something. The PM fills one out, I fill one out, and we compare shapes. The gaps between my wheel and theirs are where the useful conversation lives. I don’t use it for performance reviews or promotion cases. It’s a map of where someone is and a starting point for deciding where they want to go next.

Continue Reading »

Some companies put product marketing in product. Some put it in marketing. There’s a pattern—consumer companies lean product, enterprise companies lean marketing—but the reasoning behind it usually goes unstated.

Here’s how I think about it.

Product marketing is about information transfer. Once you accept that, the org question gets easier.

Continue Reading »

AI is making output cheap. That’s the headline.

Anyone can generate competent code, decent copy, solid designs. The bar for execution gets higher because everyone can reach it. When execution is easy, what you execute on becomes the differentiator.

The value has shifted to inputs. Higher quality input creates higher quality output. Garbage in, garbage out still applies—but now "garbage" means "the same prompt everyone else is using."

Taste Becomes the Moat

For building software, the valuable skill is knowing WHAT to build. Making trade-off decisions. Having taste.

If anyone can build anything, the people who know what’s worth building win. That’s not something AI gives you. That’s judgment. Experience. Understanding of a problem space that goes beyond what you can prompt for.

This is why "vibe coding" only gets you so far. You can tell AI to build you an app. But knowing which app to build, what features matter, what trade-offs are acceptable to your users—that requires understanding your users better than AI can.

Continue Reading »

Claude and ChatGPT each have the concept of Projects—a collection of related chats with an instruction that guides each chat. I’ve been using this as a way of learning new topics, using an instruction that tells the LLM how to design and deliver lessons to me.

Each day, I open up the chat and ask for the next lesson, and the LLM writes me a custom 10 minute read that adds to my knowledge. So far I’ve used this to learn a new statistical technique, improve my pasta-making skills, and refresh topics I’ve read in business books.

Continue Reading »

Few people are ever trained in interviewing future hires. You’re handed someone’s resume and tossed into a meeting with the candidate. The hiring manager might tell you what topics they want you to cover.

So you think back to the questions people asked you and ask versions of those. What are your strengths and weaknesses? Tell me about your greatest accomplishment? How do you prioritize?

But these don’t really tell you anything about the candidate. They don’t give you great signal about their relative skill in doing the job. What if their greatest accomplishment has nothing to do with building products? And even if it does, what are you going to learn from them describing it?

When you ask them how they do something, you’ll get an idealized version of what they would do, instead of a realistic description of how they really work.

The best way to improve your questions is to define your intent up front. Identify the traits and skills you’re looking for, and think of a way the candidate can demonstrate that skill in an interview. This usually involves asking them to describe real-world situations that used the skill.

For all your questions, write down what you are trying to learn by asking it. Think through what a strong answer would be and how you’d distinguish from a weak answer.

During the interview, probe their answers to find out what they specifically did. Product managers accomplish almost all their work through the team, so it can be hard to get them to say what they actually did. Getting this is crucial.

To get you started, here are some examples of questions I use in actual product management interviews.

Continue Reading »

I use AI as a strategic thinking partner. Not to make decisions for me—that would be abdicating leadership. I use it to expose my own assumptions and clarify my thinking.

The Dialogue Approach

The process: I frame a business challenge. The AI asks questions and makes assumptions. I clarify what I meant. This back-and-forth exposes assumptions I didn’t know I was making.

Last month, I was struggling with a product roadmap decision. I thought I had considered all angles, but through this dialogue, I discovered I was unconsciously prioritizing technical elegance over customer adoption speed. This wasn’t a deliberate choice—it was an unexamined assumption the conversation helped surface.

Continue Reading »