Every few years leadership invents a new word for an old sin. Right now that word is "founder mode." Paul Graham's essay — inspired by Airbnb's Brian Chesky, who later noted he never actually used the term — handed every executive a permission slip to skip delegation and call it conviction.
The data doesn't say what you think
The pro case is real. Bain found founder-led companies have delivered 2.1x the total shareholder returns of peers since 2015. Purdue researchers found founder-CEOs file 31% more citation-weighted patents and make bolder bets. Conviction compounds.
But the premium has an expiration date. Across more than 2,000 public companies, the founder advantage dwindles to zero roughly three years after IPO — after which founder-CEOs actively detract from value. Noam Wasserman's research is harsher still: half of founders are no longer CEO by year three, and about four in five departures are forced. Founder mode isn't a strategy. It's a phase with a shelf life.
Classify the door, not the leader
The useful question isn't "am I a founder-mode or manager-mode leader?" It's Bezos's: is this a one-way door or a two-way door?
One-way door — irreversible and identity-defining: pricing architecture, a platform bet, who owns the P&L. Get involved. Conviction is the asset here.
Two-way door — reversible and iterable: a vendor, a campaign, a feature that ships and learns. If you're in these details, you're not being rigorous. You're a bottleneck with a title.
The test takes a minute: if this goes wrong, how expensive is it to reverse? Cheap and quick means delegate. Can't-undo means you belong in the room.
Stop cosplaying
Most leaders I watch misuse founder mode to do their team's work rather than their own decisions. Skip-levels should surface information, not redo the PM's spreadsheet. The goal was never a persona — it's matching involvement to irreversibility, one decision at a time.