Strong Opinions, Loosely Held

Marty Cagan revisits the foundational product-management ideas he publicly argued for over the past two decades and identifies ten significant mistakes or blind spots. He concludes that his earlier emphasis on engineering-led product work, problem validation, and tidy frameworks underweighted business viability, the real drivers of product failure, politics, competition, and the discipline of thinking. He argues that AI has made many of his core principles more relevant, since cheaper delivery makes strategy, discovery, and build-to-learn more important than output and predictability. The piece matters because it models intellectual humility and gives product leaders a candid map of where common practice diverges from good practice.

01Key takeaways

• Treat business viability, including costs, monetization, liability, and ethics, as core product knowledge rather than a secondary concern.
• Spend more time on solution discovery, since most product failures stem from weak solutions rather than unimportant problems.
• Ask why users keep using or abandon your product, and talk to churned customers to learn it.
• Be explicit about what you do not know, and treat disproven ideas as learning rather than personal failure.
• Use roadmaps and PRDs cautiously, since they can create false confidence and crowd out outcome focus.
• Engage with corporate governance and politics, since strong product work alone may not protect the company's direction.

02Key sections

Business viability and problem discovery

Cagan admits he understated business viability, which he now sees as essential for AI products. He also argues teams overinvest in debating whether a problem is 'most important' and underinvest in solution discovery, where products usually fail.

Asking the right 'why' and staying humble

The more useful question is why customers use or abandon a product, which teams rarely ask. Cagan frames humility as knowing what you cannot know and treating wrong guesses as learning rather than failure.

Predictability, politics, and leadership

Demand for predictable roadmaps and PRDs encourages false certainty and output focus, while politics is an unavoidable part of organizations. He also admits his early work neglected product leadership, which empowered teams require more of, not less.

Governance, competition, and thinking

Great products do not guarantee great companies, so product leaders should engage with corporate governance. Competition is a full-contact sport, and process and frameworks often replace the thinking that product work requires.

Conclusion: why the product model matters more now

AI lowers delivery costs, making the futility of output-driven work clearer and build-to-learn easier. Cagan concludes that product strategy and discovery are more relevant than ever.

03From the post

“Note: This is the narrative version of a keynote talk for the Lenny & Friends Summit. Recently I was asked a question that I haven’t been able to get out of my head: “In light of all that has changed, what are the things that you used to argue were true, that you no longer... The post Strong Opinions, Loosely…”

04Frameworks mentioned

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