Things I Believe (in Product Management)

29 June 2026

• Showing beats telling. Bring a working prototype to the stakeholder meeting, not a roadmap slide.
• You don't have a problem worth solving until you can describe it in the customer's words, not your own.
• AI doesn't fail to fix bad direction, it disguises it. A bad idea built in an afternoon looks as polished as a good one.
• Strategy is what you choose to be bad at. Be good at everything and you're remarkable at nothing.
• AI de-risks the prototype, not the product. A working demo isn't a system that scales, and customers now expect both for less effort.
• Consensus is often just everyone settling quietly. Someone has to care enough to make the unpopular call.
• When everyone can ship an MVP in a weekend, the MVP stops being the differentiator. Conviction is.
• Give a competent team the wrong metric and they'll hit it perfectly. That's the failure mode no dashboard warns you about.
• As AI commoditises execution, the PMs who built their reputation on shipping fast are about to find out that was the easy part.
• Falling in love with the problem is advice. Falling out of love with your own solution is the actual skill.