Sequoia's Andrew Reed on Figma's community flywheel, the AI disruption thesis, and early investment lore

Summary

Andrew Reed, a partner at Sequoia Capital, joined the broadcast live at Figma's Config conference, where he reflected on the firm's early investment in Figma and what it took to build a community that now fills a 9,000-person venue.

The Figma investment

Reed led Sequoia's Series C into Figma in January 2019, when he was still an associate. The deal came with a memorable closing maneuver: partner Doug Leone moved a pitch meeting from the MIT conference room to the Brown conference room — because Dylan Field and co-founder Evan Wallace both attended Brown — then produced a printout showing the full history of Brown's LP commitments to Sequoia and what the firm had returned. Reed describes it as the moment he understood what elite relationship-building looks like in practice.

On the long gestation before that: Figma spent roughly four years building before it had meaningful revenue, pioneering the use of WebGL for browser-based collaborative design. Reed pushes back on treating that as a template. The time it takes a company to reach its first million dollars in revenue is basically uncorrelated with how fast it grows after that, he argues. What matters is the $1M-to-$10M trajectory. He uses OpenAI as the clearest recent example — a nonprofit for nearly a decade, then a billion in revenue within months.

The community flywheel

Reed's sharpest observation is that Config's crowd isn't a marketing outcome — it's a product outcome. The 9,000 people in the room traveled from Southeast Asia, Europe, and Africa on their own initiative. Reed draws the contrast with what most founders mean when they say "community": a dinner, free food, a mailing list. Real community, he argues, starts at the product level and is ignited by founders doing genuine customer development. Field was doing exactly that while Figma was still a developing product — not grinding in silence, but actively recruiting early users one by one.

He also credits Field's early reputation in the design ecosystem, citing Soleo Quervo, the first product designer at Facebook, who told Reed years before the Sequoia deal that Field was going to be exceptional.

AI as disruptive, not sustaining

Reed is direct that the thesis positioning AI as a sustaining innovation for incumbents is, at this point, mostly wrong. His reasoning is structural: small teams equipped with AI can now compress what used to require separate designers, developers, marketers, legal, and content functions into a single person or a handful of people. Incumbents have strong models and products, but organizational complexity prevents them from surfacing and shipping them. He uses Google's NotebookLM as the example — a genuinely impressive product that the market has already moved past, with competitors likely already replicating it in the App Store.

On the question of where AI hits first versus where prediction models expected: Reed acknowledges being consistently surprised by the order of breakthroughs. The self-driving trucks wave didn't displace blue-collar workers as feared; radiologists largely still have their jobs. Image generation emerged from diffusion research at a specific moment because of the creativity of a small group — not because of a planned roadmap. He expects that pattern to continue, with spreadsheet-driven predictions consistently lagging the actual arc.

Underwriting in the current M&A environment

On whether OpenAI's acquisition of Windsurf signals that wrapper M&A changes how Sequoia underwrites early bets: Reed is skeptical. The gestation period for the companies Sequoia targets is long enough that today's regulatory or M&A environment doesn't meaningfully shift investment criteria. The firm is still looking for what he calls "one-of-ones" — future aircraft carriers, not acqui-hire candidates. Portfolio companies may adjust whether they're buying or positioning to sell, but the underwriting standard at entry doesn't move with the news cycle.