Don't Rely On Confidence Alone

Teresa Torres argues that confidence in a product idea reflects how coherent a story feels in our heads, not whether the idea is true. She points to Research in Motion and AOL as examples of confident strategies that failed. Her remedy is to balance conviction with doubt, surfacing assumptions and asking how each could be tested. She reminds product leaders that they cannot predict the future, customers, or engineering timelines reliably. Instead of heavy upfront planning, she recommends iterative building, customer collaboration, and experiments that validate ideas before committing to them. The piece matters because it reframes product decisions as hypotheses rather than certainties.

01Key takeaways

• Treat every product idea as a hypothesis, not a settled fact, however persuasive it feels.
• Surface key assumptions explicitly and ask how you could verify each one before building.
• Replace large upfront requirement documents with iterative delivery that assumes you may be wrong.
• Run multiple small experiments and build only what proves to work instead of shipping many features blindly.
• Launch when the product is ready and coordinate marketing with customer success rather than fixed code-completion dates.
• Engage customers as collaborators, since they often know more about their needs than the team does.

02Key sections

The Trap of Confidence

Confidence comes from how neatly a story hangs together, which is different from being correct. Failed examples like Research in Motion and AOL show coherent strategies can still be wrong.

Balancing Knowing and Doubting

Drawing on Karl Weick's idea of wisdom as a balance between knowledge and doubt, Torres urges leaders to hold conviction while staying open to being wrong.

Accepting Hard Truths

Predictions, customer behavior, and engineering timelines are all unpredictable. Research and domain expertise still matter, but they should be paired with humility about what can be known.

Making Decisions Differently

Torres contrasts big upfront requirements and fixed launch dates with iterative building, collaborative discovery, and experiments that filter ideas before full investment.

Adopting an Experimental Mindset

She challenges readers to examine how many recent changes were actually tested, and notes that talking about lean methods is far easier than practicing them.

03From the post

“You find yourself in a conference room with your product team. The white board is covered with sketches. You have an idea that everyone is excited about. You mapped it out. You have a plan. Your engineering team is anxious to get started. This is what makes building products fun”

04Frameworks mentioned

Summary and takeaways written by PM Atlas; quotes are short excerpts. © the original author.