Kirby Winfield | Ascend

Winfield is a four-time operator (two IPOs, two acquisitions) who spent two years as investor-in-residence at the Allen Institute for AI before launching Ascend in 2019. That background drives a blunt thesis most polished VCs won't say out loud: pre-seed money from finance-trained "quants" is dangerous, because the people writing first checks need to understand the zero-to-one journey, not underwrite growth. He frames his own value as narrow and purpose-built for the first check, and tells founders that if they don't want to take his call again after the first meeting, they shouldn't take his money. His sharpest point is counterintuitive: traction can be a negative signal at pre-seed. He wants to bet before a founder has quit their job or built a deck, because the strongest founders won't still be raising six months later. So $200K in revenue eighteen months in reads as evidence the company lacks exponential upside, not as de-risking. Pair this with his view on talent: founders should optimize for one extreme spike rather than being uniformly good. He diagnoses his own founder career as B-plus at everything and insanely good at nothing, which is why his companies sold but never went parabolic. He credits Jake Storm at Felicis for the "spikes" language. On mechanics he gets specific. The forwardable email should lead with a LinkedIn link, because most investors check your credibility there before reading a word of your blurb, and a stale profile can sink you before the pitch starts. The deck is a leave-behind, not a performance: send it ahead, keep it near 12 slides, and never read it aloud (he says he has never once gotten excited to invest after a founder walked him through slides). On closing, he endorses tranching as the most effective catalyst short of a tier-one term sheet, and dismisses the worry that early investors resent later, higher-priced ones, calling time-based valuation fairness a fiction.