Nov 6, 2025 · 1h 18m · news
Is a $4.5BN Exit Enough in VC? & Harvey Raises $150M & Why Google is a Buy and Amazon is a Sell · 20VC with Harry Stebbings
0:00 / 0:00
▶ Watch on YouTube →
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
A panel discussion exploring the shifting dynamics of modern venture capital economics, the massive capital expenditures driving the AI revolution, and the critical transition from passive copilot tools to autonomous AI agents that are restructuring the software landscape.
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 10.4% of the talking time here. How this is scored →
Harry as informed peer 4.3 Guest teaching 3.9 Guest disagreement 3.0 Harry pushing back 2.6
speaking balance: gold is Harry, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 57:26 Jason's ultimatum to SaaS founders
Jason forcefully rejects founder excuses, arguing that anyone who hasn't reaccelerated growth via AI over the last 18 months should fire half their team or fire themselves.
Hardest push from Harry ▶ 17:35 Harry pushes back on exit sufficiencyHarry directly refuses the guests' initial framing that a $4.5B exit is automatically a great outcome, insisting that for funds over $1.5B it fails to move the return needle.
Biggest teaching moment ▶ 36:30 Rory reframes Sam Altman's capex dodgeRory dismantles Sam Altman's public reframe, educating the audience on why funding a trillion dollars in capex is a macroeconomic necessity that requires real answers rather than glib dismissals.
Harry holds his own ▶ 20:52 Harry drops insider metrics on HarveyHarry showcases deep access and expertise by citing unreleased insider metrics on Harvey's ARR, daily usage, and net retention figures.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | Harry as informed peer | Guest teaching | Guest disagreement | Harry pushing back | Why |
|---|---|---|---|---|---|---|
| Episode Trailer and Highlight Hook | 4 | 4 | 3 | 3 | Harry guides the discussion on Navan's IPO metrics and pushes back when Jason and Rory debate whether a $4.5B exit is satisfactory for large VCs. Rory educates on the realities of lockup periods, 18-month value tracking, and the mathematical differences between early and late-stage fund returns. | |
| Harvey's $150M Raise and AI's Impact on Legal Software | 6 | 3 | 2 | 2 | Harry demonstrates high domain insight by bringing exclusive insider metrics regarding Harvey's ARR, DAU/MAU ratios, and retention numbers. Rory breaks down the legal software TAM and explains the software-versus-labor spend transition required to justify an $8B valuation. | |
| VC Ownership Targets and YC's "Three on Thirty" Model | 5 | 4 | 2 | 3 | Harry cites Benchmark's McCall deal and Roger Ehrenberg's contrarian strategy to question how VCs should handle declining ownership targets. Rory explains the spectrum of capital efficiency in AI companies that has disrupted traditional ownership rules. | |
| Sam Altman's Public Spat and the Trillion-Dollar AI Capex Question | 5 | 2 | 4 | 4 | Harry aligns strongly with Jason on board dynamics, criticizing board members who fail their fiduciary duty out of fear of damaging founder relationships. Rory delivers a sharp critique of Sam Altman's dismissive public response regarding OpenAI's trillion-dollar capex commitments. | |
| Big Tech's AI Earnings and Capex: Google, Amazon, and Meta | 4 | 4 | 3 | 2 | Harry probes the guests on Big Tech's cloud growth and valuation shifts across Amazon, Google, and Meta. Jason dismisses Amazon's public GPU arrangement with OpenAI as theatrical positioning rather than structural leadership. | |
| SaaS Giants and the Bounded Tech Valuation Universe | 3 | 5 | 6 | 2 | Jason takes a contrarian, aggressive stance demanding that SaaS founders fire half their teams if they failed to achieve AI-driven growth reacceleration in the last 18 months. Rory agrees with the core thesis, explaining the stark valuation gap between AI-relevant SaaS and private equity turnarounds. | |
| B2B SaaS and Genuinely Replacing Humans with Agents | 3 | 5 | 3 | 2 | Harry acknowledges Jason's past assertions coming true as the market shifts from copilots to full human replacement. Rory outlines the concrete economics of replacing $40k employee salaries with $10k annual software agents. | |
| Open Evidence and the Dynamics of Rapid Market Saturation | 5 | 4 | 2 | 2 | Harry brings data comparing Open Evidence's rapid adoption curve directly against Doximity's historic timeline. Rory analyzes the risks of rapid S-curve adoption, noting that fast individual adoption can quickly lead to market saturation without broader expansion products. | |
| The Series A Investing Landscape and Agentic Re-architecture | 4 | 4 | 3 | 3 | Harry prompts the guests on whether Series A investing is harder than ever. Jason rejects the premise, contending that an unprecedented explosion in seed supply makes Series A top-of-funnel the best it has ever been. | |
| Stiffening VC Competition and Lessons from Loss Stories | 4 | 4 | 2 | 3 | Harry opens up about losing a recent competitive deal to Andreessen Horowitz, prompting Rory to detail how mega-cap firms moving earlier has elevated competitive pressure. The group ends with a forced-choice comparison between Kalshi and Polymarket before engaging in light banter. |
Statements from this episode (39)
Lemkin Refuses to Take Meetings With 'Mortal Founders'
“For me, it's like, I don't even want to take meetings with mortal founders.”
O'Driscoll: Fully Optimized Founder Fundraising Causes VCs to Miss Ownership Targets
“A founder's optimized fundraising is a VC's below ownership target.”
O'Driscoll: Navan's IPO Proves Bill Gurley's Pricing Critique Is Wrong
“Actually, this actually shows that sometimes Bill Gurley is wrong because his idea is that these IPO share allocations are called free money. You get your IPO. It always goes up and the leaving value on the table. This is an example of an IPO where, you know, …”
O'Driscoll: True VC IPO returns are best measured 18 months post-IPO
“When we used to track exits every year, which we do, we used to track as of IPO, and then we used to call it the locked in value, 18 months later. And our mental model was that when you want to figure out how much money people actually made, look at the market…”
O'Driscoll: Mature 30%-growth SaaS companies are valued at 6-7x NTM revenue
“Your operating assumption should be that for mature companies, SaaS transaction type, whatever, right? It's not just SaaS. It can also be non-recurring revenue companies with decent margin profile and 30% growth. We are back to six to seven times NTM, right?”
Lemkin: Recently closed an early-stage deal at a $50M post-money valuation
“I just did a deal with some founders I love and it was very expensive for me. I did a deal at 50 post.”
O'Driscoll: Tech companies now need $400M-$500M revenue to go public
“You now have to assume that that 400,000,500 million is the threshold for an IPO.”
Stebbings: Lightspeed turned $257M into $1B on Navan for ~4x return
“Lightspeed was two 57 to a billion, so it's just under a four X blended.”
O'Driscoll: Late-stage VC funds cannot return a fund from a single deal
“You're not going to get a late stage investment court returning the fund. It's not a thing because if you're doing 20 deals by And evenly that's, you know, what's that? Five percent each, unless you get a 20 X, which you typically don't get in the late stage d…”
Stebbings: AI startup Harvey raised $150M at an $8B valuation
“Harvey raises a hundred and fifty million bucks Eight billion dollar valuation led by our dear friends at Andreessen.”
Stebbings: Harvey hit $150M ARR with 170% net dollar retention
“A little insider at Harvey one of that not so quiet investors shared with me that they're at a hundred and fifty million in ARR. Their DAU to MAU ratio is 40%, which I thought was astounding. Like, usage-wise, phenomenal. GRR, 98%. NDR, a 170%.”
Lemkin: Harvey's valuation implies $400M in forward annual recurring revenue
“If it's 400 forward revenue, that's 20 X. Agreed. That's what it is. 400 next year. They're, 400 ARR. That's what they're predicting. 400 ARR. Not GAP. 400 ARR. That's my guess.”
O'Driscoll: Selling software to lawyers is a bad business
“Selling software to lawyers is a particularly shitty business. It's a more constrained business. You've got to do more. You've got to help more. You've got to speed them up”
Lemkin says his last three VC investments yielded 6-8% ownership
“And the last three investments I've done are in the six to eight percent range, even though that's my rule.”
Scale Venture Partners targets 10-11% average equity ownership per deal
“It's hard with our stage, you know, our target would be 10 to 11% on average, you know, probably late A or B, some earlier, some later than that.”
Lemkin: YC Advises Founders to Sell Max 10% Equity Around Demo Day
“Y Combinator's advice for most companies is to raise a maximum of 10% before, during, and after demo day.”
Lemkin: Anthropic projects $70 billion in revenue by 2028
“Anthropic is now projecting seventy billion in revenue in 2028”
Lemkin: VCs refrain from giving critical feedback to successful founders
“I think there is so much fear among VCs of getting out of step with the most successful founders. There's so much fear and you guys are going to disagree with me, but I see it all across my portfolio. The better the company is doing, the more everyone's a grin…”
Stebbings: VC board members abandon fiduciary duties to protect founder NPS
“Board members will not intervene in any way to protect the shareholders because of the bad NPS that will come from damaging that founder relationship. That, to me, is one of the most egregious escapes from fiduciary duty, and this is some of the most reputable…”
O'Driscoll: AWS lost cloud dominance by failing to adapt to AI
“Five years ago, AWS dominated cloud compute. Now they don't, right? Because cloud compute evolved from being simple compute to being AI-centric computing, and they didn't evolve fast enough with it.”
Lemkin: Google is underappreciated while Amazon is overappreciated
“I do think that for the moment, Google is underappreciated and Amazon is overappreciated.”
O'Driscoll: Meta spending $70 billion on AI is inefficient
“Doesn't appear to be the most efficient way to spend seventy billion dollars, as judged from the amount of infighting, but we'll see.”
O'Driscoll: Legacy SaaS giants have lost high-valuation growth premium
“These companies, they're in a much more bounded world now. They might grow 15%. They might grow five percent. They might trade at six times revenues. They might trade at four times revenues. You know, if you look at the stock chart, the area of magic is gone, …”
Rory O'Driscoll: Palantir is cash flow positive trading at 123x revenue
“Palantir is trading at 123 times revenues. And this cash flow positive.”
Lemkin: Fire half your team if AI hasn't reaccelerated growth
“If you haven't gotten a boost this year from AI, fire half your team right before the holidays, give them a turkey and three months of severance, but they failed. You have a, your team is not good enough. They had 18 months to ship a product that mattered in t…”
Lemkin: HubSpot and Salesforce must demonstrate AI revenue growth in 2026
“HubSpot and Salesforce got to deliver in 20, 26 because they're in play. They have the AI things they built, they ship the products, but they haven't seen the bump that data dog and Twilio and Mongo have”
Lemkin: Salesforce Agentforce is a legitimate, competitive product
“There's 2000 people building agent force and we've deployed it. It's pretty good. It's quite good. It's very competitive with any other agent you're going to buy.”
Lemkin: Harvey will improve next year and replace more legal associates
“Whatever Harvey is today, it's going to get better. And I guarantee you it's a better piece of software next year than it is at eight billion. Right. And it will replace more mediocre associates”
O'Driscoll: Companies can replace $40k employees with $10k AI agents
“There are areas where right here, right now, you can replace a 40,000 dollar worker with a 10,000 dollar a year agent and be better off.”
O'Driscoll: Salesforce must automate 20-30% of customers within two years
“In the next two years, they got to get 20% of their customer base plus maybe 30% onto being as pointy edge as Jason in terms of automation. Otherwise, these guys are going to go somewhere else to get their automation.”
Stebbings: Open Evidence Reached 300,000 Doctors 10x Faster Than Doximity
“Open evidence grew to 300,000 doctors in one year, which is one 10th of the amount of time it took Doximity.”
O'Driscoll: Vertical AI Market Size Equals Professionals Times Work Automated
“You should know for each of the tools, professional does, how many of that profession exist, how many lawyers, how many doctors, how many bankers, how many wealth advisors, cause that's your time, you know, and I'm tracking that. And then it, so it's really nu…”
O'Driscoll: Enterprise AI Adoption Will Take 5 to 6 Years
“The pace of corporate adoption is much slower. So you're exactly right. I don't think everyone's going to buy agent force other equivalent in a year. It might be five years, not 10 like SAS. It might be half the time to SAS, but it still could be five or six y…”
Jason Lemkin: Series A is experiencing its best-ever deal funnel
“This is a, the best of times to be a Series A investor because there is just an explosion of seed AI startups. There's so many and they can't all get funded. And I know it's stressful. I know it's hard, but it is a gift that thousands of founders are in San Fr…”
Jason Lemkin: Seed stage venture investing is currently at its worst
“And I think it's the worst for seed because Everyone each year wants to be a seed investor even more. It's the, everyone, you know, what, you know, Jake Paul, Jake Logan, whatever, which are the Jake's you know, it's not just the chain smokers anymore, and Jar…”
Rory O'Driscoll: Enterprise software over next decade will shift to agentic AI
“I do at least feel now since ChatGPT that the direction of architectural direction in enterprise B to B for the next 10 years Is pretty much obvious in a given, you know, it's some form of agentic software. I hate that word. It conveys a lot, but some form of …”
O'Driscoll: Large VC Funds Are Expanding Into Revenue Series A and B
“Given the fund size, all the large firms that were, you know, typically C to A are doing all those deals. So the competition set has stiffened, right? You're up against tougher, better firms, and you've got to bring up your A game, right?”
Lemkin: Kalshi is a safer long-term bet than Polymarket
“So based on my limited knowledge, I'm going Kalshi because I feel like it is a safer long-term bet than someone that is writing the current political vibes, which are all in favor of everything here. It just could, it was a lot different a couple of years ago,…”
O'Driscoll: Sports leagues, not political administrations, will drive prediction market regulation
“Just to make a prediction and which is in theme in keeping with the idea here. I predict that any re-regulation won't happen because of any new administration. I think the real challenge to sports betting like this will actually be the leagues themselves wrest…”