ExpertsFM20VC: Inside Sequoia's Investment Committee: Lessons from Don Valentine, Doug Leone and Alfred Lin | How the SpaceX and Citadel Deals Went Down | What Sequoia Specifically Looks for in Founders with J
The Twenty Minute VC (20VC): Venture Capital | Startup Funding | The Pitch · August 24, 2026
5 Selected Lessons:
heuristic · 4 case · 1
Founders-you-like vs founders-who-make-money matrix
Source
The Twenty Minute VC (20VC): Venture Capital | Startup Funding | The Pitch — 20VC: Inside Sequoia's Investment Committee: Lessons from Don Valentine, Doug Leone and Alfred Lin | How the SpaceX and Citadel Deals Went Down | What Sequoia Specifically Looks for in Founders with J (August 24, 2026) · at 0:00 ▸ · at 33:15 ▸
When this applies
when deciding whether to partner with or bet on a person you personally dislike or find abrasive
The takeaway
Separate your personal affinity for someone from your assessment of their capacity to produce the outcome you actually need.
Why it works
Likability and money-making ability are two independent axes; collapsing them into one judgment causes you to pass on profitable but unlikable people or back likable but unprofitable ones.
One level up
Decisions should be organized around the variable that actually predicts the outcome, not the variable that is easiest to feel.
Where it stops holding
This stops working when the trait you dislike (e.g. arrogance) is actually masking a real weakness rather than being the cost of a genuine spike.
Detail
rule: Plot founders on a 2x2 of 'founders you like' vs 'founders who make you money' and figure out which quadrant you're actually in before deciding.
threshold: If there's no underlying 'spike' (unique strength) behind a disliked trait, treat it as a red flag rather than an acceptable cost.
rationale: Many of the best investments come from founders partners didn't personally warm to, so liking someone is not the signal that predicts returns.
In their words
If you look at founders you like versus founders who make money as a two by two matrix, your job is to figure out in which part of the quadrant we make money.
Ask for the worst reference, not just the best
Source
The Twenty Minute VC (20VC): Venture Capital | Startup Funding | The Pitch — 20VC: Inside Sequoia's Investment Committee: Lessons from Don Valentine, Doug Leone and Alfred Lin | How the SpaceX and Citadel Deals Went Down | What Sequoia Specifically Looks for in Founders with J (August 24, 2026) · at 40:43 ▸
When this applies
when evaluating anyone who controls which references/testimonials you get to see
The takeaway
Deliberately ask for the negative/worst case data point, not just the favorable one, because the selection of favorable evidence hides the information you actually need.
Why it works
Best references are self-selected to be favorable and reveal little; asking for the worst reference tests honesty directly (how they answer, not just what they say) and surfaces genuinely useful texture about weaknesses.
One level up
The informativeness of a data point is inversely related to how much the subject controls its selection.
Where it stops holding
This stops working when the person refuses to name a worst reference at all or names a trivial one, in which case the question itself becomes the signal (evasiveness) rather than the reference's content.
Detail
rule: After a founder names their best reference, follow up by asking who their worst reference would be and why.
threshold: n/a
rationale: Not looking for perfection, looking for clarity — how directly and honestly someone names and explains their worst reference is itself valuable signal, and the subsequent reference call gives real texture.
In their words
As they finish, you ask the counter question, which is, who would be your worst reference and why. And see their calls changed. People answer that honestly.
Humility rule: 'we are only as good as our next investment'
Source
The Twenty Minute VC (20VC): Venture Capital | Startup Funding | The Pitch — 20VC: Inside Sequoia's Investment Committee: Lessons from Don Valentine, Doug Leone and Alfred Lin | How the SpaceX and Citadel Deals Went Down | What Sequoia Specifically Looks for in Founders with J (August 24, 2026) · at 0:00 ▸ · at 40:43 ▸
When this applies
when a team or individual's past success risks reducing the intensity of preparation for the next decision
The takeaway
Deliberately track and surface your misses, not just your wins, to keep humility and preparation intensity calibrated against the true difficulty of the task.
Why it works
Explicitly counting the companies you missed (every public company Sequoia has 'seen at some point' but didn't back) rather than only the ones you won keeps the base rate of failure visible, preventing overconfidence from past wins.
One level up
Survivorship-bias correction (counting the misses, not just the hits) is a discipline that sustains performance over time.
Where it stops holding
This stops working when excessive focus on misses becomes paralyzing rather than motivating, undermining the confidence needed to act decisively.
Detail
rule: Remind the team that for every company that went public that you backed, there are far more that you saw and missed — so bring full energy and preparation to every meeting regardless of past success.
threshold: n/a
rationale: New Sequoia hires write this sentence on their first day as an intimidating, humbling anchor.
In their words
We are only as good as our next investment.
Overestimate the next year, underestimate the next decade
Source
The Twenty Minute VC (20VC): Venture Capital | Startup Funding | The Pitch — 20VC: Inside Sequoia's Investment Committee: Lessons from Don Valentine, Doug Leone and Alfred Lin | How the SpaceX and Citadel Deals Went Down | What Sequoia Specifically Looks for in Founders with J (August 24, 2026) · at 50:04 ▸
When this applies
when forecasting adoption speed of any disruptive new technology or behavior change
The takeaway
Calibrate forecasts by discounting near-term (1-year) exuberance and inflating long-term (10-year) expectations, since both are typically mis-estimated in opposite directions.
Why it works
People consistently extrapolate near-term change too aggressively (because recent momentum is salient) while failing to compound long-term structural shifts (because exponential change is hard to intuit), producing systematically wrong short vs long horizon forecasts.
One level up
Human intuition is linear, so predictions about exponential processes need a deliberate correction applied to both short and long horizons.
Where it stops holding
This stops working when the technology/behavior in question is actually adopted unusually fast due to unique conditions (e.g. zero switching cost, no incumbent process to displace), making even the 1-year estimate too conservative.
Detail
rule: Always overestimate what happens in a year and underestimate what happens in 10.
threshold: n/a
rationale: Enterprises won't move as fast as hype suggests short-term, but structural shifts (like agents as a new customer) compound more than expected over a decade.
In their words
I'm always reminded, always overestimate what happens in a year and underestimate what happens in 10.
Airbnb: controversial seed bet that became one of the highest money-on-money returns
Source
The Twenty Minute VC (20VC): Venture Capital | Startup Funding | The Pitch — 20VC: Inside Sequoia's Investment Committee: Lessons from Don Valentine, Doug Leone and Alfred Lin | How the SpaceX and Citadel Deals Went Down | What Sequoia Specifically Looks for in Founders with J (August 24, 2026) · at 18:20 ▸
When this applies
when the surface description of an idea sounds absurd or low-status but the underlying mechanics may be transformative
The takeaway
The biggest return multiples tend to come from bets that are genuinely controversial at the time, not from consensus deals — and conviction has to be sustained across subsequent rounds, not just the first check.
Why it works
Sequoia led a seed round that every other firm had passed on; the willingness to back a seemingly silly idea, combined with sustained conviction through later rounds (not just the initial bet), produced one of the fund's best-ever returns.
One level up
Outsized returns require tolerating social/reputational discomfort that consensus-chasing competitors won't accept.
Where it stops holding
This stops working when the 'controversial' idea is controversial because it's actually a bad business, not because it's merely ahead of consensus understanding.
Detail
situation: Airbnb's founding idea (air mattresses, strangers' floors) sounded like a bad idea and had been rejected by most VCs.
decision: Sequoia led the seed round and kept investing in subsequent rounds rather than treating the seed as a one-off bet.
outcome: One of the highest money-on-money returns in Sequoia's history.
lesson: Controversial-sounding ideas with sustained follow-on conviction, rather than consensus deals, are where outsized small-to-big dollar conversions happen.
In their words
I think where you turn small dollars into big dollars, they have to be controversial.