Instincts vs Ideas

Two things founders conflate

Mark Pincus draws a distinction between an instinct and an idea, and argues that most founders fuse the two. An instinct, in his usage, is an underlying truth about human behavior, the emotional charge a person feels at a broken experience. An idea is a specific product implementation of that instinct. His claim is that instincts tend to be right and durable, while ideas often need to be discarded and rebuilt. Pincus presents this as the single core lesson of the product-management course he taught at Stanford, the one point he hoped students would internalize above all else.¹

The sharp edge of the framework is his own line: "A B+ and a B+ is the enemy of an A. B+ often, if we stick with the B+, is taking up space. It's distracting us from unlocking the A. And to fail or to call the ball that that idea isn't that good, it in itself is really powerful and not many people do it."¹ Once a founder calls that ball, they are free to look around: whether someone else is pursuing the instinct better, whether the world is not yet ready, or whether four other versions are worth trying.¹

Pincus is careful about what improves with experience. He does not claim his instincts became more accurate over time. What improved, in his own words, was calibration: "my ability to calibrate, to both separate the two and recognize that idea just isn't it, that's what's gotten better."¹ He describes the feel of a correct pairing as one where everything works and the work does not feel like pushing a rock up a hill, and says that asking whether something could be it is itself a sign that it is not.¹

The cases

Pincus illustrates with his own failures. He registered smstaxi.com years before Uber, holding the correct instinct that people would order cars through their phones alongside a weaker idea, doing it over SMS, and never pursued it. He launched Tribe.net months before LinkedIn and other social networks, with working virality, and still failed, because the specific idea for what the network was for was wrong even though the instinct behind social networking was obviously correct. His own retrospective verdict is blunt: "If I could go back to Mark 2003 and shake myself: 'You're on the right path, but this one idea isn't it.'"¹ Both cases carry the same lesson in his framing: the product work can be technically right and the venture can still fail if the idea is wrong, so a right instinct is necessary but not sufficient. The trap he most wants founders to see is that a B-plus idea with just enough traction to survive becomes a cage, absorbing team, capital, and energy that would otherwise find the A. This connects to the fake CEO failure mode, where a founder who has fused their identity with an idea resists calling it out to protect team cohesion.

Where the instinct is trained

Pincus's framework concerns calibrating an instinct once a founder has one. Alex Bouaziz addresses the upstream question of where a durable instinct comes from, and his answer is customer proximity. He argues that what makes founders effective, even without a background in a field, is the volume of time they spend with customers, and reports keeping hundreds of customers on WhatsApp so that he builds instincts about what they want in a way competitors cannot replicate.² Read alongside Pincus, the two accounts complement each other: customer proximity is the training data that produces a trustable instinct, and instinct-versus-idea calibration is the discipline for acting on it. Bouaziz adds a caveat that mirrors Pincus's own view of fallibility, noting that trusting an instinct can lead a founder into a wall but at least keeps them true to their own read. The framework's open questions follow from that fallibility: distinguishing a wrong instinct from a right instinct with a wrong idea is available mostly in hindsight, and the cases usually cited are ones where the instinct turned out to be correct.