Startup bets, ranked like playoff legends

The Venture Hall of Fame

A Ringer-style ranking of the best startup bets of the last 15 years, according to Formation.

Not just the biggest outcomes. This board rewards the checks that combined fund-bending return potential, company scale, lonely conviction, and check-writing difficulty. And a little bit of personal preference.

Scorecard

100

Blended
Ranking

35%

Return

How violently the check bent the fund math.

25%

Outcome

Company scale, strategic importance, and market gravity.

25%

Non-Consensus

How lonely the take looked when the check cleared.

10%

Difficulty

Structure, competition, capital intensity, and weirdness.

Narrative / Author Discretion

The mythology bonus. The "Oooo" factor.

Top 10 board

The rankings

Ten bets, one board, zero false humility. Sort the sheet and if you have a better list send it to me on X @lmushin or leeor@formationvc.com.

Rank

2019Seed / first major VC check

OpenAI

The Weirdest Seed Check Ever Became the Most Important Company in the World

Investor

Vinod Khosla

Khosla Ventures

Ringer Comp

2012 LeBron Game 6 in Boston

Why It Was Great

Khosla backed OpenAI when it was still structurally weird, commercially unclear, and governed in a way most venture investors would have rejected immediately. It is now one of the largest private companies ever.

Why It Was Non-Consensus

Nonprofit origins, capped-profit structure, unclear monetization, massive compute needs, and no obvious path to normal venture governance.

Outcome

OpenAI became the defining AI company.

Rank

2023Series C

Anthropic

The Other Frontier AI Lab That Suddenly Wasn’t Other Anymore

Investor

Yasmin Razavi

Spark Capital

Ringer Comp

Tom Brady at Pick 199

Why It Was Great

Spark leaned into Anthropic before it was obvious that the market would support multiple OpenAI-scale frontier labs. The company quickly became one of the central pillars of the AI ecosystem. A cold ping from Yasmin started it all.

Why It Was Non-Consensus

At the time, many investors still believed OpenAI had already won, foundation model companies were too capital intensive, and AI safety-oriented governance could limit commercial velocity. It has been said 21/22 VCs passed.

Outcome

Anthropic became one of the highest-valued AI companies globally.

Rank

2018Series B

Figma

The Browser-Based Design Tool That Beat Adobe at Its Own Game

Investor

Mamoon Hamid

Kleiner Perkins

Ringer Comp

Steph Curry 2009 Draft Pick

Why It Was Great

Mamoon’s Figma Series B was a perfect venture bet: an obsessively loved product, a giant incumbent, and a market that looked too narrow until collaboration turned it into a platform.

Why It Was Non-Consensus

Design tools looked like Adobe’s kingdom. Browser-based professional creative software was not yet obvious. The market seemed smaller than the eventual outcome.

Outcome

Figma became the defining design collaboration platform and a major public-market software company.

Rank

2013Series A

Coinbase

The Bitcoin Exchange Before Bitcoin Was Respectable

Investor

Fred Wilson

Union Square Ventures

Ringer Comp

Drafting Josh Allen Before the Mechanics Were Fixed

Why It Was Great

USV backed Coinbase when crypto was fringe, reputationally risky, and institutionally radioactive. The company became the public-market gateway to the crypto economy.

Why It Was Non-Consensus

Bitcoin was associated with scams, drugs, regulatory uncertainty, and hobbyist speculation. Most mainstream investors wanted no part of it.

Outcome

Coinbase became the first major public crypto company in the U.S.

Rank

2017Seed / early

Anduril

Defense Tech Was Radioactive. Then It Became Inevitable.

Investor

Trae Stephens

Founders Fund

Ringer Comp

The 2000 Ravens Defense

Why It Was Great

Trae Stephens and Founders Fund backed/founded Anduril when defense tech was culturally unpopular in Silicon Valley and procurement looked impossible for startups.

Why It Was Non-Consensus

Most VCs hated hardware, defense, government sales, and anything associated with military procurement. Anduril was all four.

Outcome

Anduril became the flagship defense technology startup.

Rank

2016Seed

Cerebras

The Wafer-Scale Nvidia Counterpunch That Actually Made It to the Public Markets

Investor

Eric Vishria / Steve Vassallo / Pierre Lamond

Benchmark / Foundation Capital / Eclipse

Ringer Comp

Dirk Nowitzki Before Anyone Believed in Stretch Bigs

Why It Was Great

The early Cerebras investors backed an almost comically ambitious chip architecture: build a processor out of an entire silicon wafer and go straight at the most powerful company in computing. The company turned a moonshot hardware bet into a public-market-scale outcome.

Why It Was Non-Consensus

Hardware was supposed to be too slow, too capital intensive, and too dangerous for classic venture. Taking on Nvidia with a novel wafer-scale system looked like walking into traffic.

Outcome

Cerebras became one of the most important public-market challengers in accelerated computing.

Rank

2012–2013Seed / early

Canva

Australia’s Design Startup Became a Global Software Monster

Investor

Rick Baker

Blackbird

Ringer Comp

Giannis at Pick 15

Why It Was Great

Blackbird’s early Canva investment became a fund-making and firm-making position. Few venture bets have had this much institutional impact on an entire ecosystem.

Why It Was Non-Consensus

Australian consumer/prosumer design software was not the default place to look for a global decacorn. It looked too simple, too far away, and too un-enterprise.

Outcome

Canva became one of the most valuable private software companies in the world.

Rank

2013Series A / early

Databricks

The Open-Source Research Project That Became the Lakehouse Empire

Investor

Ben Horowitz

a16z

Ringer Comp

Nikola Jokic in the Second Round

Why It Was Great

Databricks turned Apache Spark from a beloved technical project into one of the most important enterprise data companies in the world. Ben Horowitz was underwriting a deeply technical founding team, an open-source wedge, and a market that did not yet have the language for the lakehouse category.

Why It Was Non-Consensus

Open-source infrastructure companies were still hard to underwrite. The company started around academic technology, competed in a brutally technical data market, and had to convince enterprises that a new data architecture could sit beside or eventually challenge the warehouse.

Outcome

Databricks became one of the defining data infrastructure companies of the cloud era.

Rank

2012Company creation

Snowflake

The Cloud Data Warehouse That Looked Suicidal Until It Was Obvious

Investor

Mike Speiser

Sutter Hill Ventures

Ringer Comp

The Spurs Drafting Manu Ginóbili at 57

Why It Was Great

Sutter Hill helped create Snowflake and backed it before cloud data warehousing was a mainstream category. The company became one of the biggest software IPOs ever.

Why It Was Non-Consensus

Competing against Amazon, Google, Microsoft, and entrenched data infrastructure vendors looked irrational. The category did not yet look inevitable.

Outcome

Snowflake became a foundational cloud data platform.

Rank

2014-2015Series A / B / B extension

DoorDash

The Food Delivery Bet Everyone Thought Was a Low-Margin Knife Fight

Investor

Alfred Lin / John Doerr

Sequoia Capital / Kleiner Perkins

Ringer Comp

J.J. Watt Before the Three DPOYs

Why It Was Great

Some investors kept backing DoorDash through a period when food delivery looked crowded, operationally ugly, and structurally bad. The company became the dominant U.S. local commerce platform.

Why It Was Non-Consensus

Food delivery had terrible unit economics, brutal local operations, intense competition, and a history of failed startups.

Outcome

DoorDash became the dominant U.S. delivery marketplace.

Honorable heartbreak

Just Missed

The bench is absurd. Sure they are sad to miss the Top 10 all the way to the bank.

Scale AI

Snub Watch

Dan Levine / Accel

Pre-AI-boom, data labeling looked like a small, services-heavy corner of the market. The better read was that high-quality data would become central infrastructure.

Rippling

Snub Watch

Mamoon Hamid / Kleiner Perkins

Parker Conrad was on a revenge mission after Zenefits, but not everyone believed he would pull it off.

SpaceX

Snub Watch

Founders Fund and others

Maybe the best private company investment ever, but much of the legendary return story predates the 10–15 year window.

Instacart

Snub Watch

Sequoia and others

Huge company, but the return narrative is messier after late-stage valuation compression.

CoreWeave

Snub Watch

Magnetar / Coatue / others

Incredible AI infrastructure timing, with an even stranger origin story: a crypto-mining pivot that became one of the defining compute bets of the AI boom.

Methodology

How the board was built

The ranking blends scoring with my personal preferences. So does being early when everyone else thought it was dumb.

35%

Return

How violently the check bent the fund math.

25%

Outcome

Company scale, strategic importance, and market gravity.

25%

Non-Consensus

How lonely the take looked when the check cleared.

10%

Difficulty

Structure, competition, capital intensity, and weirdness.

Narrative / Author Discretion

The mythology bonus. The "Oooo" factor.

Too high? Too low? Who got snubbed?

Reach out to @lmushin on X or leeor@formationvc.com.