Sep 30, 2021 · 36m · 20vc
20VC: The Opendoor Memo: Keith Rabois on The Origins of Opendoor from a Conversation with Peter Thiel, Why Cash is Not a Competitive Moat for Startups Today and What People Misunderstand About Black Swan Events in Real Estate and How it Impacts Opendoor
0:00 / 0:00
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
In this episode of 20VC's 'The Memo,' Founders Fund General Partner Keith Rabois details the incubation, scaling, and public listing of Opendoor, sharing foundational insights on algorithmic home pricing, venture building, and capital dynamics. He also reflects on applying these operational lessons to his latest venture, OpenStore, while offering strategic perspectives on venture capital moats and personal productivity routines.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. Harry holds 27.6% of the talking time here. How this is scored →
Harry as informed peer 3.1 Guest teaching 3.9 Guest disagreement 2.4 Harry pushing back 1.7
speaking balance: gold is Harry, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 9:08 Keith rejects top-down market sizing
Keith explicitly dismisses Harry's question about top-down market analysis, stating he never does top-down models and that residential real estate obviously requires no such modeling.
Hardest push from Harry ▶ 24:42 Harry pushes on cash as a moatHarry pushes Keith on whether cash functions as a moat in capital-intensive businesses, forcing Keith to directly refute the premise.
Biggest teaching moment ▶ 16:54 Keith re-educates on down-market liquidity dynamicsKeith corrects the common intuition about real estate crashes by presenting data from 2008 showing minimal 30-day price decay and explaining why liquidity providers perform better during market downturns.
Harry holds his own ▶ 28:34 Harry brings data on SPAC underperformanceHarry confronts Keith with the stat that 97% of SPACs trade below IPO price, challenging the general validity of SPACs as a public listing mechanism.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | Harry as informed peer | Guest teaching | Guest disagreement | Harry pushing back | Why |
|---|---|---|---|---|---|---|
| Welcome Bumper and Conversation Kickoff | 2 | 3 | 1 | 1 | Harry opens the show by asking Keith about the origin story of Opendoor and why the 2003 prototype was deferred. Keith explains how the post-dot-com crash capital crunch and lack of tech debt facilities prevented funding two home cohorts. | |
| Capital Dynamics and Home Pricing Cohorts | 3 | 5 | 4 | 2 | When Harry asks about performing top-down market analysis, Keith dismisses the premise, stating he never does top-down market modeling because residential real estate is obviously the largest asset class in the world. | |
| Assembling the Founding Team and Talent Assessment | 3 | 3 | 1 | 1 | Harry asks how Keith detects talent in 'diamonds in the rough' when assembling founding teams. Keith highlights matching candidate attributes directly to core startup risks and praises Eric Wu's interviewing skills. | |
| Overcoming Early Seller Conversion and Trust Hurdles | 3 | 3 | 1 | 1 | Harry inquires about early operational hurdles. Keith shares an epiphany regarding seller conversion where sellers doubted a startup's solvency, which was solved through targeted PR, before discussing city expansion criteria. | |
| Pricing Strategy Debates with Co-Founder JD Ross | 3 | 6 | 4 | 2 | Keith refutes common wisdom around black swan macro real estate events, explaining that Opendoor actually performs better in down markets due to heightened liquidity demand and backing it up with 2008 price decay data. | |
| Fundraising Challenges and Correcting Operational Errors | 3 | 4 | 3 | 1 | Harry asks about past fundraising friction. Keith candidly admits to making operational mistakes in 2016, such as building distributed local teams copied poorly from Uber, before contrasting those lessons with OpenStore. | |
| Financial Discipline Amid Back-to-Back Funding Rounds | 4 | 5 | 4 | 3 | Harry asks whether cash itself forms a competitive moat for asset-heavy models. Keith disagrees directly, asserting that cash merely follows progress rather than predicting or establishing a true moat. | |
| Expanding to 60 Markets and the SPAC Decision | 4 | 4 | 2 | 2 | Harry asks about expanding to 60 markets and the choice to go public via SPAC. Keith explains that direct listings had legal constraints on capital raising at the time, making a SPAC the fastest vehicle for public listing. | |
| Evaluating Public Listing Vehicles and SPAC Mechanics | 4 | 4 | 3 | 3 | Harry brings up the statistic that 97% of SPACs trade below their offering price. Keith rejects the broad generalization, emphasizing that different listing mechanisms fit different companies and citing Rocket Lab as a success. | |
| Personal Routine: Sleep, Barry's Bootcamp, and Fitness | 2 | 2 | 1 | 1 | In the quickfire section, Keith discusses unsung heroes at Opendoor, prioritizing eight hours of sleep, and how Kevin Systrom introduced him to Barry's Bootcamp. |
Statements from this episode (21)
Thiel offered $5M in 2003 to fund Opendoor's predecessor
“Peter was willing to fund the company with about five million dollars, but fundamentally, to buy enough homes and prove that you could accurately price the homes, five million was at the very low end of what would be feasibly, statistically, to prove that we c…”
Rabois: Debt capital is the operational oxygen for Affirm and Opendoor
“In addition, we do use debt and debt is the oxygen for businesses like Affirm. It's the oxygen for companies like Opendoor.”
YouTube and PayPal founders helped build Opendoor's 2003 predecessor
“The logo for Home Run was designed by Chad Hurley as a side project. I got a decent part of the website was coded by Max Levchin.”
Rabois: 20 to 25 homes provides statistical validity for pricing algorithms
“And so we could buy a cohort of call it 20, 25 homes. And you can get some statistical validity out of that size cohort.”
Rabois: Opendoor launched in 2013 with $10M initial capital
“Which is actually what we did in 2013 when we funded the company. We actually only funded the company with ten million dollars, which was basically just enough to buy two cohorts of homes.”
Rabois: 83% of US homes are commodities, not unique assets
“And in truth, most homes, when I say most, I mean, 83% of homes in the United States are more like a commodity.”
Rabois: Infallibility in assessing talent alone can propel a startup
“And for CEO, that one skill alone Can propel a company. If you can assess people infallibly, that is, you know, the propellant for many, many, many companies.”
Rabois: Early city expansion cannot be driven by data or algorithms
“When you're going from one city to two cities, There is no algorithm behind this stuff. I mean, because by definition, you don't have data to tune what cities are most attractive.”
Rabois: San Francisco Bay Area real estate is too anomalous for Opendoor
“The Bay Area is just too anomalous from a real estate perspective to invest in.”
Rabois: Opendoor still executes its pricing model incorrectly eight years later
“The one that I still make fun of JD about is how we did our pricing model, which we still basically don't do correctly. You know, eight years later, we'll eventually fix this, but there's a lot of debate about whether we should have consumers name their own pr…”
Rabois: Opendoor performs better in depreciating real estate markets
“I've always had the hypothesis that, in fact, Opendoor would do better in a bad real estate market, a down market, a depreciating market, than an appreciating market. If you think about it, we basically provide liquidity. Typically, people who want liquidity c…”
Rabois: Phoenix home prices decayed at most 2% monthly in 2008
“No market in the United States actually collapsed very fast. Like even in Phoenix, which was one of the most affected markets in 2008, the greatest decay in any 30 day period was about two percent.”
Rabois: Diversified US residential real estate lost no money in 2008
“If you studied residential real estate in the United States over the last century, had you been relatively well diversified across all markets in the US, you would not have lost money even in 2008. So most US markets do not move in tandem or Historically have …”
Rabois: Copying Uber's distributed local market team model is a bad idea
“Second, and related to that, we built insanely dumb idea of building like a distributed team, like in local markets, which is the bad idea taken from Uber was never a good idea in the first place. It might have even been a bad idea for Uber, but it was a bad i…”
Square reduced merchant onboarding from weeks to instant approval for 93%
“Before Square, people sometimes forget, before Square in the United States to get access to the credit card infrastructure. You can apply, in quotes, for a merchant account And that would take three weeks. It would require a credit check and a manual investiga…”
Rabois: Opendoor negotiated its second funding round before launching
“So we wound up negotiating the second round pretty much before we actually launched the company.”
Rabois: Cash is not a competitive moat for technology startups
“I don't really believe cash is a moat generally across technology companies. I think cash follows progress, not really predicates it.”
Rabois: Opendoor will cover all US markets except possibly Bay Area
“So literally everywhere in the United States are going to be able to, you know, buy and sell homes through Opendoor except maybe the Bay Area.”
Rabois: Opendoor pursued a SPAC under the misconception it was faster
“The real reason for choosing to go down the SPAC model was I wanted us to be a public company as fast as humanly possible. I had some reasons to believe we could do it a little bit more rapidly through a SPAC. The truth is that's not really true, but it was ea…”