Jason Lemkin, Sam Lessin and Frank Rotman Debate the Future of Seed Investing

Harry Stebbings asks Sam Lessin why he believes the institutional seed-investing model faces more than a brief downturn. Lessin describes the prior era as a startup factory: seed funds prepared companies to meet the next round's metrics, passed them through successive investors and expected a predictable public-market exit. He argues that many such companies failed to justify those expectations and that the factory model will not return in the same form.

Lessin says seed investing will instead be more bespoke and governed by a power law: a few exceptional successes and many losses. Jason Lemkin questions whether SaaS and cloud are fundamentally broken or simply need a reset to conditions before the 2019–21 boom. Frank Rotman agrees that businesses should be de-risked in stages, but argues that cheap, abundant capital encouraged companies to fund too many initiatives before building a solid core.

Rotman proposes a more disciplined progression from seed through later rounds, with real failure between stages rather than automatic markups. His suggested 20–30% failure between stages is a framework for the discussion, not a measured forecast for every cohort. In his view, a company must learn enough and advance far enough to justify the next financing.

Lemkin recalls helping early investments such as Pipedrive, Algolia and Talkdesk hire and present themselves to Series A investors. Lessin calls that work akin to investment banking and says his firm also does it. But he argues that the most important returns in his own portfolio came from unusual, inexpensive ideas that were difficult to package for the prevailing playbook.

Lessin expects mortality to rise and says chasing an obvious, highly competitive deal at a high price leaves little room for exceptional returns. Rotman adds that investors had been trying to earn the right to do what everyone agreed was attractive. The participants discuss their own funds and investment histories, so those examples reflect their financial interests.

Stebbings says he does not see seed prices correcting: multistage funds and wealthy technology executives are still competing. Rotman sees the correction moving gradually from public markets toward early stages, with insider rounds buying companies time. He warns that a high seed price can lead later investors to decline to bid if the business has not grown into the valuation.

Lemkin doubts founders will accept this risk when they believe strongly in their companies. Lessin acknowledges that even he has sometimes invested personally in expensive deals because the absolute check was small. He says his firm now tends to pass unless it sees substantial upside, rather than relying on a predictable downstream markup.

Rotman cautions that thousandfold outcomes usually require an exceptionally large market and that a standard dilution percentage is not a substitute for a valuation thesis. He argues that many companies can produce good returns only by using capital efficiently and earning money at a relatively small scale.

Lessin points to businesses started with little or no outside capital, including Muck Rack and a business founded by his wife, as evidence that profitability can create options. He would like more seed investors to support businesses that grow through revenue, with further fundraising as a choice rather than an obligation. These examples are his account, not a guarantee that every company can follow that path.

Asked about exits of a few hundred million dollars, Lessin says a properly priced, capital-efficient seed investment can make such an outcome attractive. He thinks many viable businesses have markets too small for a public listing but can still become valuable. Rotman agrees that the question is how much a team can learn and de-risk for each dollar invested.

Lessin challenges the assumption that proving one or two hypotheses will always unlock an efficient Series A market. He would rather see seed capital used to establish a working business that can approach profitability, leaving later capital optional. His view is a proposed strategy, not a claim that every startup can become profitable immediately.

Lemkin raises Trade Desk as an example of the classic seed outcome, using a hypothetical in which early funds retain 20% each to a roughly $40 billion valuation. He explicitly notes that investors distributed earlier; the resulting $8 billion calculation is an illustration, not their actual realized proceeds. He asks whether comparable ownership and outcomes are still available today.

Lessin cites his firm's Solana seed investment and Teamshares as examples of unusual opportunities that others did not initially favor. He describes a 1,000-fold result for Solana and says an early Teamshares check purchased a substantial stake. These are his own investment accounts, with no independent verification or implication that similar returns are typical.

The group debates fund size. Lessin argues that very large seed funds make it difficult for even a major win to generate a standout fund multiple. Stebbings counters that a $50 million fund may struggle to lead enough $3–4 million seed rounds while preserving diversification and reserves. Lessin replies that concentrated selection is part of the challenge.

Rotman describes a Crunchbase analysis suggesting only a small single-digit percentage of venture-backed companies in historical vintages went public, with his examples ranging around 2–4%. He treats an IPO or large sale as a marker for the outcomes many venture funds seek, while acknowledging that a $300 million sale after a small investment can also yield an excellent return. His analysis and its scope are not independently verified.

Lessin asks whether the industry must organize itself around those rare IPO-scale outcomes. He favors companies that can become profitable and worth $100–300 million while retaining the option to grow much larger. Lemkin questions whether downside protection is the best use of a venture investor's time compared with finding the next Trade Desk.

Lessin argues that building a good, efficient company first need not forfeit the possibility of an extraordinary outcome later. Rotman agrees that a good business can become a platform for a great one, or reveal that selling is the better choice. He emphasizes that attractive smaller exits require discipline in how much capital has been invested.

Rotman illustrates the economics with a hypothetical ten-to-one enterprise-value-to-paid-in-capital ratio. Depending on ownership and financing, he suggests earlier investors might earn more than later investors. A $300 million outcome would require less than $30 million of invested capital to meet that illustrative ratio; actual investor proceeds depend on terms and ownership.

Lemkin worries that some overfunded founders no longer respect the value of investor capital. He recounts asking one unnamed founder to return part of an investment and receiving a dismissive response. It is his account of one interaction, not an independently verified characterization of founders generally.

Lessin is more optimistic that founders will see the appeal of owning a larger share of a profitable business. He argues that many entrepreneurs may prefer control and flexibility to a highly diluted stake in a large venture-backed company. His comparison is a forecast about choices, not a claim that one path is always superior.

Lemkin counters that a software company without outside capital can fall behind well-funded rivals once it reaches a certain scale. He acknowledges the excesses of venture funding but argues that some categories need investment to compete. Lessin responds that there are many industries and approaches beyond the SaaS race; they do not agree on one model for all founders.

Rotman offers another decision framework: each dollar invested should produce evidence for or against the business thesis. When go-to-market works and the model improves, capital may accelerate momentum. When acquiring growth becomes more expensive and margins deteriorate, more revenue can conceal “anti-proof”; simply applying the same revenue multiple may encourage additional investment in a weakening model.

The discussion turns to Y Combinator and party rounds. Lessin describes YC's playbook and demo day as the clearest expression of the factory model, saying standardized packaging makes it harder for him to identify a distinctive signal. He distinguishes that critique from the continued availability of money from wealthy angels. These are his opinions about the model, not findings about every YC company.

Rotman acknowledges YC's brand and role in generating deal flow, but worries that some companies are encouraged to prioritize non-scalable early growth for demo day. He says investors may later need to change that approach. Lemkin takes a different view: he values YC as an imperfect filter for drive and ability among first-time, outsider founders.

Lemkin says his difficulty is often ownership rather than price or founder quality. He cites his investments in Algolia and RevenueCat, but says today's fragmented rounds can make it hard for his fund to buy enough of a company to justify the work. His comments reflect his investment strategy and interests.

Rotman prefers a cap table with active investors who offer complementary contacts and skills rather than dozens of passive checks. Asked whether Series A investors care if a seed round was a party round, he says no. Lemkin adds that a group of operators may still be useful to a founder if they respond thoughtfully to specific questions.

Lessin provocatively says his standard answer when asked what support his firm gives founders is “nothing,” while acknowledging that it does provide help. He wants founders who can execute without depending on a fund. A founder who prefers many small investors and can manage independently may reasonably choose that structure, in his view.

Rotman distinguishes needing an investor to run the business from benefiting from specialist help. He points to QED's credit and financing experience as useful to a capable fintech founder working through credit policy or debt funding. Lessin accepts that different seed strategies can work; his own emphasizes unusual ideas others have overlooked.

Lemkin argues that recruiting senior leaders is an area where even talented first-time founders can benefit from real assistance. He criticizes venture talent teams as often failing to deliver the support they advertise. Lessin says early-stage sales hiring is especially difficult and that someone who consistently finds excellent early sales leaders could create substantial value; their criticisms are opinions based on their experience.

Stebbings revisits a wager: Lemkin expects roughly one IPO a week in the second half of 2024. Lemkin says he sees efficient, growing companies with more than $200 million in revenue in his own portfolio and believes a window could open if market multiples hold. He explicitly acknowledges that his timing may be too optimistic.

Lessin takes the other side, arguing provocatively that public investors may prefer established technology giants to smaller companies with $200 million in revenue. He offers investment in large AI-exposed companies as his view of the alternative, not as personalized financial advice. He questions whether merely good businesses will command the valuations once expected.

Rotman thinks Lemkin's direction may be right but his timing too early. He sees companies reshaping their plans for efficiency and believes public investors can value sustainable, growing businesses, while betting against one IPO each week in that period. All three views are forecasts made in 2023, not reports of what later happened.

Lemkin argues that an IPO can be priced to give public investors room for long-term gains and cites earlier software listings as possibilities. He recalls the 2008–09 crisis and a faster-than-expected recovery to explain his optimism. Lessin instead sees a future more concentrated among very large companies and private-equity-style combinations.

Rotman says QED had five portfolio companies go public in 2021 and sees additional candidates across its fintech holdings. He extrapolates that other sectors may also have a substantial pipeline, while acknowledging that it is not yet clear which startups will make it. The figures and prospects are his account of a portfolio in which his firm has an interest.

Stebbings suggests large firms may continue to raise funds despite weaker returns. Lessin disagrees, pointing to reports of reduced fund targets and combined vehicles, while allowing that leading firms can still raise capital. Lemkin cites his historical estimate that the average SaaS IPO had absorbed roughly $390 million before listing and predicts capital will return if IPOs revive; that estimate is not a current market statistic.

Rotman says conversations during a recent QED fundraise suggested LPs were consolidating commitments and scrutinizing middling funds and emerging managers. He expects the adjustment to play out over years rather than instantly. He also says a reopening IPO market could change those conditions if liquidity returned to LPs.

In the quick-fire round, Rotman highlights the return of capital efficiency and expects founders to need more concrete results after a narrative-led seed round. Lessin says he wants to fund unusual companies at low valuations that can become profitable without relying on Series A investors. Lemkin says he would prefer more concentrated rounds that let him acquire at least 10% of companies he chooses.

Stebbings thanks the guests and jokes about the exposure from their opposing IPO bets. The panel closes after a discussion in which the investors have defended distinct strategies rather than reached a single seed-market formula.