On November 4, Sequoia Capital posted a memo on X titled "The Sequoia of Tomorrow." Almost every item in it is a retreat: it is sunsetting Arc, the accelerator it launched in 2022, and Scouts, the program it started around 2009; closing the London office it opened in 2020; merging its early and growth teams into one; scrapping the Monday partner meeting; and handing ownership and operation of the firm to seven general partners (GPs). The same memo says Sequoia has raised roughly $10 billion in new funds (memo text and personnel changes per TechCrunch, which reported on it and reprinted it in full the same day).
The two programs being shut are exactly the two routes that brought outsiders closest to Sequoia. Scouts gave founders and operators small checks to invest on the firm's behalf; Jason Calacanis used it to bring Uber to Sequoia, and Sam Altman brought Stripe. Arc was Sequoia's answer to Y Combinator: several seven-week cohorts a year, with admitted companies getting $500,000 to $1 million up front.
Sequoia's own framing is "putting more energy behind fewer people and things." The deal record in the SVTR AI Venture Database shows that shift began a year ago.
One year after the handoff, the money moved later
In November 2025, Roelof Botha stepped down and Alfred Lin and Pat Grady took over as co-stewards (see our profile of Alfred Lin at the time). Splitting at that date, here are the AI rounds Sequoia joined in the 12 months before and after:
Table 1 | AI rounds joined by Sequoia Capital, 12 months before and after its November 2025 leadership change
| Metric | Nov 2024 – Oct 2025 | Nov 2025 – Nov 6, 2026 | Full database, same periods |
|---|---|---|---|
| Rounds joined | 69 | 101 | 2,346 → 4,097 |
| Early-stage rounds | 25 | 28 | 1,439 → 2,485 |
| Growth and late-stage rounds | 29 | 53 | 455 → 948 |
| Early-stage share of rounds with known stage | 46% | 35% | 76% → 72% |
| Median round size (USD millions) | 50 | 100 | 14 → 21 |
Data: SVTR AI venture database, as of 2026-11-06. "Joined" covers Sequoia Capital and the Sequoia Scout Fund. Early stage means angel, pre-seed, seed and Series A; growth means Series B and C; late means Series D and beyond. Strategic investments, secondary sales and rounds with undisclosed stage are excluded from the share. Round size is the total round, not Sequoia's own check.
In one year, the growth and late-stage rounds Sequoia joined rose from 29 to 53, while early-stage rounds went only from 25 to 28. The median round size doubled from $50 million to $100 million. The whole database also shifted later over the same period, but far less: the early-stage share fell 4 percentage points across the database and 11 points at Sequoia.
Shutting Arc and Scouts writes into the org chart what had already happened in the deal record.
London closes, but European deals did not shrink
The reason given for closing London is "We simply don't see the need for a local presence." Our data does not support reading this as "there are no deals in Europe": in the 12 months after the handoff Sequoia joined 12 rounds in European companies, up from 8 in the prior 12 months. They include:
- French accounting software company Pennylane, Series D
- German modular industrial robotics company RobCo, Series C
- UK reinforcement-learning company Ineffable Intelligence, seed
- Flow Engineering, which builds AI agents for hardware engineering, Series B
Sequoia wants Europe's deals, not a European office. Luciana Lixandru, hired from Accel in 2020 as Sequoia's first partner in Europe, declined a move to San Francisco. "Europe is where my family is, and where I want to build," she wrote on X. She remains a partner, keeps her board seats, and will invest personally in European founders.
Sequoia's early door is narrowing
Our view: Sequoia's early-stage entry point has shrunk from a wide net to the individual judgment of seven GPs. Teams that are pre-seed and outside the Bay Area will find Sequoia money harder to get; the ones that get in will mostly be companies a GP went out and chased.
Peers make this clearer. In the 12 months after the handoff, counting Sequoia Capital alone, it joined 26 early-stage AI rounds. Over the same period a16z joined 77, General Catalyst 62, Lightspeed 50, and even the now-separate HongShan joined 36. By number of early rounds, Sequoia is no longer in the top tier.
In August we wrote about Sequoia's single-sponsor rule: if one partner is willing to stake their own standing on a deal, it can go ahead. With fewer doors, that rule only matters more.
The numbers may carry a bias: large growth rounds get more press, small early rounds are easier to miss, so Sequoia's early share may be understated. But under the same method the database as a whole fell only 4 points, and the extra 7 points at Sequoia are hard to blame entirely on missed deals.
Three things to watch
- Early rounds Sequoia joins per half-year. Our database shows 15 in the first half of 2025, 9 in the second half, and 14 in the first half of 2026. If the first half of 2027 drops below 10, the early-stage pullback is real; if it holds around 15 while growth keeps rising, Sequoia is reserving early deals for GPs to do personally rather than stepping back.
- Who picks up the early-stage ground. Stephanie Zhan, who joined Sequoia in 2015, announced last month that she is leaving to start her own firm. Watch her first fund size and the stage of her first ten deals: if most are seed, someone is already filling the space Sequoia left.
- European pace without a London office. Sequoia joined 12 European rounds in the 12 months after the handoff. If that number falls clearly in 2027, "supporting Europe from the Bay Area" was just a phrase; if it holds, European founders will have to get used to dealing directly with partners in the Bay Area.
Every Sequoia deal, with stage, amount and co-investors, can be looked up round by round on the Sequoia Capital profile in the SVTR AI Venture Database.