Justin Fishner-Wolfson | Sourcery Summary

Guest: Molly O'Shea, Co-Founder and Managing Partner, 137 Ventures Host: Justin Fishner-Wolfson, Sourcery

1. Key Themes

The "Companies Stay Private Longer" Thesis — And How It Created an Entire Asset Class

137 Ventures was built on a single structural insight in 2011: that the Jobs Act removing the 500-shareholder rule would eliminate the primary forcing mechanism pushing companies to go public. That insight has since been validated at scale.

"Our belief was companies were going to stay private longer... And I think our view was very simply that companies were going to end up staying private longer. And what that meant was there would be a lot more opportunities to invest in those businesses." 00:02:56

By 2025, secondaries represented $240 billion and 31% of all venture volume — a market that expanded by "a couple orders of magnitude" since the firm launched.

Position Building Through Concentration, Not Diversification

137 Ventures made approximately two dozen investments in SpaceX alone across 16 years, and has written four to five checks into Hadrian. This is fundamentally different from the VC norm of writing one large check, hitting an ownership target, and stopping.

"We want to keep investing in the companies that we have conviction in and really get concentration in those businesses. And I think that's a little bit different than how the industry works, right? Most people are trying to write like one large check at the beginning." 00:17:51

Business Model Disruption as the Core SpaceX Moat

The shift from cost-plus government contracting to firm fixed-price contracts — combined with partial rocket reusability — was identified as the foundational economic breakthrough, not just the engineering.

"They flipped the business model in the industry. Because historically it had been cost plus and they had transitioned to a firm fixed price model. And so a combination of a business model shift and then making the Falcon 9 partially usable fundamentally changed the economics of the business." 00:07:06

Starlink's Inflection Point Was Identifiable — If You Were Paying Attention

The thesis on Starlink becoming a massive business became clear around 2019 when the constellation economics proved out. The TAM framing was unusually precise.

"Who are all the people who don't currently have access to high-speed internet? Right? I mean in the US it's like tens of millions of people. And like that's in the US. Like forget globally, right? Like there are tens of millions of people who don't have — not that they can't afford it — they literally just can't buy it." 00:11:35

Companies, Not Categories — The Anti-Thematic Investing Philosophy

137 Ventures explicitly rejects sector-based investing in favor of identifying the single dominant company within a space, and concentrating there rather than diversifying across competitors.

"If you looked at SpaceX you would think, oh gosh, there are all these great space investments. There were not. We looked at probably every space investment for 15, 20 years and the good investment was SpaceX... there's a category but really there's a company that's incredibly successful and that you should continue to invest in that company." 00:15:18

The Coming IPO Wave — SpaceX Starts a Multi-Quarter Public Market Event

OpenAI and Anthropic have both filed S-1s confidentially, and SpaceX is positioned as the bellwether that opens the floodgates of long-delayed private company liquidity.

"Companies from a life cycle perspective are starting to go public again. SpaceX is the big one to kind of start this. But I mean as you mentioned like OpenAI and Anthropic are both — they've filed their S-1s confidentially. So it's going to be a pretty interesting few quarters in the public markets." 00:00:18

Ownership Percentage Is the Wrong Metric — Cash on Cash Returns Are What Matter

Molly explicitly calls out ownership percentage as a heuristic that causes investors to pass on great deals, referencing Peter Thiel's willingness to take 10% of Facebook when others demanded more.

"I look back at what Peter did with Facebook. You know I think a lot of the reason that he originally did that deal was because most people wanted a much higher ownership percentage. And he was okay getting 10% of the company which obviously was clearly the right decision." 00:23:29

Tender Process Design as a Competitive Moat and Talent Tool

Managing tender logistics — who is included, at what frequency, and with what expectations — is itself a sophisticated discipline that 137 Ventures has developed into a structural advantage.

"A lot of this is just like incentive alignment and like expectations management, right? If you run a tender it turns out your employees are gonna assume that you're gonna run another tender, right. And for the most part people don't really like it when prices go down." 00:18:50

The "Launch and Learn" Philosophy as a Cultural Export from SpaceX

The test-to-learn ethos that made SpaceX successful has spread to the broader aerospace and deep tech ecosystem, and Molly frames public test "failures" as the correct methodology — not bad news.

"It's not a failure as long as you're learning something... you can't sit in a clean room or run a model that will predict all of the possible issues. And so as long as you're getting enough data from the process it's incredibly valuable and I think SpaceX has done it incredibly well." 00:27:22

2. Contrarian Perspectives

Secondaries Are Not a Niche Strategy — They Are the Industry

When 137 Ventures launched, institutional LPs did not believe the secondary market would become significant. 16 years later, it is 31% of all venture volume. The contrarian bet was obvious on the ground but invisible from institutions.

"When we went out and talked to institutional limited partners, they were not necessarily of the mind that the industry was changing, that this was going to be a thing. And so at this point we more or less won the argument. But that was not true 16 years ago." 00:05:48

Foundational AI Model Investing Is Unknowable — And 137 Ventures Has Passed

While the consensus VC move has been to pile into foundation model companies, Molly articulates a principled reason for abstaining: you cannot identify which model will be dominant in 10 years, and that uncertainty disqualifies investment.

"We haven't invested in any of the big foundational models. I think it was just always sort of hard for us to understand which one was going to be the best model... which one is going to be the one that is the best in 10 years? And I don't know how to answer that question." 00:16:32

SpaceX's Equity for Blue-Collar Workers Is an Underappreciated Social Event

The fact that SpaceX distributed meaningful equity down to welders and electricians — not just software engineers — means the IPO will be genuinely life-changing for a class of workers who have never participated in tech wealth creation before.

"I think it's great that they've really shared equity across the entire company, which is incredible, and to your point they have lots of people on the shop floor who literally make all of the stuff that matters. So the fact that this is life changing for so many people I really think is incredible." 00:22:48

Ownership Percentage Is a Distraction That Costs Investors Enormous Returns

The entire industry has anchored on ownership targets as a discipline, when in fact adjusting the expected exit size makes ownership percentage irrelevant — and fixating on it caused most investors to miss Facebook at the seed stage.

"You don't really want to get dogmatic on the things that don't matter. You want to stay focused on things that do matter, which is just what are the returns to the funds, to your LPs. And as long as you're hitting that cost of capital you absolutely want to invest and like getting to an argument about ownership percentage is sort of distracting." 00:23:59

Facebook's Early Secondary Market Policy Created Perverse Incentives — a Warning for Today's Private Giants

Facebook's right-of-first-refusal approach functionally forced employees to quit in order to sell shares — a cautionary governance lesson directly applicable to today's massive private companies managing tender programs.

"They basically started telling people like if you sell shares, we're gonna fire you, right? And so people just started quitting, right? Because they wanted to sell. And they're like, so they created this very perverse incentive of like okay, well if you're a current employee you have to become a former employee to sell." 00:19:46

3. Companies Identified

SpaceX

Reusable rocket and satellite internet company. The central investment thesis of the episode — 137 Ventures has made approximately two dozen investments since 2008, first accessed via Founders Fund. Cited for the business model shift from cost-plus to firm fixed-price, reusable Falcon 9, launch frequency approaching 200 per year, Starlink constellation, and broad equity distribution to all employees. IPO is imminent and described as the catalyst for the broader private-to-public liquidity wave.

"They've just built things on top of that over the years like Starlink, which has been amazing." 00:07:35

Palantir

Data analytics and government software company. Cited as potentially 137 Ventures' very first portfolio investment, predating even their SpaceX positions at the new firm.

"I actually think our first investment was in Palantir. I might have to go check with that. But I actually think our first investment was actually in Palantir." 00:13:41

Cognition

AI coding and software agent company. Identified as uniquely positioned as the last independent AI agent company, giving enterprise customers access to all underlying foundation models without lock-in — analogous to multi-cloud strategies.

"Cognition kind of sits in a really important part of the market right now where if you're an enterprise customer, it's like these are the only guys who can give you access to all the underlying models without the lock-in that comes from that." 00:17:27

Hadrian

Precision manufacturing company for aerospace and defense. Cited as a portfolio company where 137 Ventures has written four to five checks — an example of their concentration strategy in action.

"Whether or not it's Hadrian, we've probably written, I don't know, four or maybe five checks there at this point, right? Like we want to keep investing in the companies that we have conviction in." 00:17:51

Impulse Space

In-space propulsion and last-mile orbital delivery company. Founded by Tom Mueller (SpaceX's first propulsion engineer). Cited for unique capabilities in commercial and government markets, and as an example of 137 Ventures backing SpaceX alumni.

"We got excited about what they were doing and kind of the unique capabilities that they had and how that was specifically relevant for commercial but also for government as well." 00:14:15

Anduril

Defense technology company. Named as a major portfolio position and cited alongside SpaceX as an example of a single dominant company within a category — perhaps a slightly less extreme version.

"Anduril might be like maybe a little less extreme version of that. But it's just like there's a category but really there's a company that's incredibly successful and that you should continue to invest in that company." 00:15:18

Varda Space Industries

In-space manufacturing company. Mentioned in the context of founder Will Bruey, who previously worked at SpaceX and sold shares in an early SpaceX tender.

"The best story is actually like from Will Bruey because he's the founder at Varda but he was also at SpaceX and so he sold in a tender a very long time ago." 00:04:49

OpenAI

AI foundation model company. Cited as having filed a confidential S-1, positioning it alongside Anthropic as part of the imminent public market wave.

"OpenAI and Anthropic are both — they've filed their S-1s confidentially. So it's going to be a pretty interesting few quarters in the public markets." 00:14:38

Anthropic

AI safety and foundation model company. Cited alongside OpenAI as having filed a confidential S-1.

"OpenAI and Anthropic are both valid — they're S-1s confidentially." 00:00:18

Uber

Ride-sharing company. Named as a major 137 Ventures portfolio position. No further elaboration provided.

Gusto

Payroll and HR platform for small businesses. Named as a major 137 Ventures portfolio position. No further elaboration provided.

Ramp

Corporate spend management and finance automation platform. Named as a major 137 Ventures portfolio position. No further elaboration provided.

Facebook (Meta)

Cited as the historical case study for why companies were forced public (500-shareholder rule), why that pressure disappeared (Jobs Act), and as Peter Thiel's pivotal early investment that validated flexible ownership percentage thinking.

"I look back at what Peter did with Facebook... he was okay getting 10% of the company which obviously was clearly the right decision." 00:23:59

4. People Identified

Molly O'Shea

Co-Founder and Managing Partner of 137 Ventures, $15 billion AUM. Former Founders Fund investor who led the 2008 SpaceX investment — the first outside institutional capital in that company. Built a firm around the secondary market thesis before it was accepted, making roughly two dozen investments in SpaceX alone.

"Back at Founders Fund we did the deal in 2008. That was sort of the beginning. Because at that point Elon had mostly funded it himself and we were really the first outside institutional capital." 00:00:18

Elon Musk

Founder and CEO of SpaceX. Cited for a first-principles approach to problem solving and the ability to update beliefs rapidly when new data arrives.

"Elon is probably like a — it's just a first principles approach to everything and the flexibility that when you have new data you can easily change your mind. That's the core of it." 00:28:36

Gwynne Shotwell

President and COO of SpaceX. Cited for her composure under pressure — described as the calm, stabilizing presence during the Falcon 1 Falcon third launch post-failure debrief. Her fix for the stage separation issue was reduced to a single insight: add a delay between stages.

"I think I was standing next to Gwynne or something and she — I was like, oh, you know what, so how are you guys gonna fix this? And she was just like, I think we should add a delay between the stage separation." 00:26:14

Tom Mueller

Co-Founder of Impulse Space, former SpaceX VP of Propulsion (employee #1). Called out as the go-to person for moving things in space. His immediate recognition of Starlink's potential was noted.

"Tom's the guy you want to use if you're trying to get things to move in space. Right, if you want to put up a constellation, Tom's a good guy to talk to." 00:13:19

Will Bruey

Founder of Varda Space Industries and former SpaceX employee. Cited as an example of the human reality behind tender programs — sold SpaceX shares in an early tender and used proceeds for everyday life events.

"The best story is actually like from Will Bruey because he's the founder at Varda but he was also at SpaceX and so he sold in a tender a very long time ago." 00:04:49

Peter Thiel

Co-Founder of Founders Fund and PayPal. Cited as the model for flexible ownership-percentage thinking — took 10% of Facebook when other investors demanded higher stakes, which Molly holds up as the canonical example of cash-on-cash return focus over ownership dogma.

"I look back at what Peter did with Facebook. I think a lot of the reason that he originally did that deal was because most people wanted a much higher ownership percentage. And he was okay getting 10% of the company which obviously was clearly the right decision." 00:23:59

Brett Johnsson

CFO of SpaceX. Cited as a relatively recent addition to SpaceX leadership (15-16 years tenure) and highlighted as someone to watch as SpaceX navigates its public company transition and earnings calls.

"We're gonna get to a bunch of earnings calls with Brett. It's gonna be fun seeing him as a public company CFO." 00:29:11

Glenn (Shotwell context — likely Glenn referenced separately)

A long-tenured SpaceX team member cited for 22 years at the company and described alongside Elon as one of the key leadership lessons. Identified for unshakeable calm and problem-solving ability under stress.

"Glenn really just doesn't — there's like no problem that you just can't work, right? It's just like no matter how stressed everyone is, it's like if you can be the calm person in the middle who can just help people get to the right answer — that's incredibly valuable." 00:28:36

Eric Romo

13th employee of Impulse Space. Cited for publicly predicting 137 Ventures would become a household name in the venture world.

"Eric Romo said — if you don't know 137 Ventures, you will know it very soon." 00:01:16

5. Operating Insights

Design Tender Programs With Long-Term Expectations in Mind Before You Run the First One

Once you run a tender, employees immediately assume the next one is coming. The decisions made in the first tender — who is included, at what price, with what frequency — set expectations that are very hard to reverse. The Facebook case shows what happens when a company improvises this under pressure.

"If you run a tender it turns out your employees are gonna assume that you're gonna run another tender, right. And for the most part people don't really like it when prices go down. And so you want to think about: how frequently are you gonna do these things? Who are you gonna allow? Are you gonna allow current employees, former employees?" 00:18:50

Include Current Employees in Liquidity Programs — Excluding Them Creates Perverse Churn Incentives

The Facebook case is the canonical failure: employees who wanted to sell had to quit first because the policy only allowed former employees to transact. The lesson for operators running fast-growing private companies is to proactively design current employee inclusion from the start.

"I would argue you probably want to include your current employees and you know maybe you want to include your former employees. But if you're gonna build this over time maybe you don't include them on the first one. You include them over time, right? So a lot of it is just kind of expectations management and building a good process." 00:20:15

The Right Heuristic for Iterative Hardware Development Is: Are You Getting Data?

For deep tech and hardware companies, a test "failure" is only a failure if you learned nothing. The SpaceX third Falcon 1 launch — where stage separation timing was the entire problem — was solved in one conversation by adding a 7.5-second delay. That is the standard to hold test programs to.

"It's not a failure as long as you're learning something... you can't sit in a clean room or run a model that will predict all of the possible issues. And so as long as you're getting enough data from the process it's incredibly valuable." 00:27:22

6. Overlooked Insights

Cognition May Be the Most Strategically Undervalued AI Company Because of Multi-Model Neutrality

This was mentioned briefly and without emphasis, but the insight is significant: enterprise customers are replicating the multi-cloud playbook in AI — deliberately avoiding lock-in to any single foundation model. Cognition, as the only remaining independent AI agent company, is the sole provider of model-agnostic enterprise access. This is not a features story; it is an infrastructure moat story.

"Enterprise customers really don't want to get locked into a single foundational model in the same way that when you sort of thought about cloud — it was like people really weren't sure whether or not they wanted to be locked into a single cloud provider. So people were like multi-tenant... Cognition kind of sits in a really important part of the market right now where if you're an enterprise customer, it's like these are the only guys who can give you access to all the underlying models without the lock-in that comes from that." 00:16:58

SpaceX's ~200 Annual Launches Create a Compounding Frequency Moat That Is Entirely Separate From Cost

The launch frequency figure — approximately 200 per year versus 15-20 for anyone else globally — was mentioned in passing but represents a structural compounding advantage that goes beyond economics. More launches mean more data, more reliability records, more customer trust, and faster iteration cycles. No competitor can close this gap without first closing the frequency gap, which requires years of operational history they do not have.

"They were launching not just cheaper but way more frequently, right? They're gonna launch, I don't know what it's gonna be this year, but something close to 200 times. Like no one else in the world is launching even 15, 20 times. Right? So what they've accomplished has really not only increased the access to space because it's cheaper but because you can actually get there faster." 00:08:24