San Francisco —  The pitch deck was unremarkable. The market was contested. The founder spent the first eight minutes of the meeting disagreeing with our partner about a thesis we'd published two weeks earlier. We wired the term sheet that Friday.

Every venture firm has a private list — the founders we backed despite ourselves. They were rude in diligence, dismissive of feedback, allergic to the playbook. Several of them are now running the most valuable companies in their cohort. The uncomfortable question, the one we don't put in the LP letter, is whether we backed them because of the friction or in spite of it.

What conviction actually looks like

"Coachability" became a checklist item the same way "TAM" did — a heuristic that survived past the moment it stopped being useful. The best founders are not uncoachable. They are selectively coachable. They absorb data voraciously, and they reject opinions ruthlessly, and they can tell the difference between the two in real time. That is not a personality flaw. That is the job.

"The founders I'm proudest of having backed all argued with me on the first call. The ones I regret were the ones who agreed with everything."

I. The first-meeting tell.

Three of the partners we interviewed for this piece described a near-identical signal: a founder who answered a question with a sharper version of the question. Not deflection — refinement. It indicates the founder has lived inside the problem long enough to know which framings are wrong.

II. The reference call paradox.

Reference calls on the strongest founders are almost always mixed. Half the references talk about brilliance and energy and pace. The other half use the word "difficult" without elaborating. Investors who only back the founders whose references are uniformly glowing systematically miss the outliers.

III. What the firm actually offers.

The honest version of the value prop, when you back this kind of founder, is not advice. It is cover — capital, credibility, and the patience to let a strong opinion compound for eighteen months without quarterly second- guessing. Firms that cannot articulate this internally tend to either avoid the archetype entirely or fund it and then quietly try to manage it.

The bet under the bet

Backing an uncoachable founder is a bet that the disagreement that makes them hard to work with is the same disagreement that produces the company. Most of the time it isn't. When it is, the returns are not normally distributed. They are the firm.

Camilla Verdoso writes about founders, conviction, and the venture firms backing both. Reporting based on interviews with three early-stage partners on background.