Is the Venture Capital Model Broken? A Debate on Prices, Product-Market Fit and Returns

Harry Stebbings opens with a question about seed funds facing $30 million post-money valuations. Eric Paley says the model is not dead: some hot companies command prices many investors reject, but he sees much of the market moving back toward more rational levels. For a founder, an overpriced seed can make the next round harder when the company is already priced at what a Series A investor might pay.

Mike Maples resists sweeping macro claims, but says seed returns still depend on backing a non-consensus founder who turns out to be right. He points to historically modest entry prices for businesses such as Lyft, Airbnb and Dropbox. A fashionable deal priced for perfection leaves less room for returns and may signal an idea that is popular rather than distinctive.

Jason Lemkin asks what can still be non-consensus when so many software categories are established. Maples responds that an active investor must find something the market does not yet understand; otherwise an index would be a simpler choice. The difficult work is preparing the mind to recognize an entrepreneur's insight before it becomes obvious.

Paley describes his firm's analysis of the hottest investment theme in each of the previous 20 years. He says the most valuable company founded in each year, measured at the time of the conversation, was never from that year's hottest theme. He recalls OpenAI's founding year, when direct-to-consumer commerce was fashionable, as an example; his recollection of the exact year is tentative.

Stebbings asks how contrarian ideas fit seed funds when infrastructure, climate or biotechnology may require large amounts of capital. Maples argues that founders create movements around such opportunities and seed investors can take risks before the rest of the market. But he distinguishes that from paying a high price for a currently popular AI model company, where a thousandfold return may be implausible.

Paley says the incentives can pull investors toward consensus: early markups help them raise the next, sometimes larger, fund. Paper value relative to paid-in capital, or TVPI, is not the same as cash distributed to investors, or DPI. He expects a substantial gap between the two for years, while Maples calls patience a form of arbitrage.

Stebbings asks whether their reasoning rules out seed investments in AI. Lemkin says the next financing round matters and, in his 2023 conversations with B2B investors, companies without an AI story may struggle to raise a Series A. Paley counters that evidence from successful experiments can attract capital without an enormous markup, though funding might be harder.

Paley distinguishes using AI as another part of software from chasing a company solely because it has an AI narrative. His approach is to follow founders with a compelling map of the future, including founders working in AI, rather than invest thematically because a category is popular. He warns that short-term incentives can inflate a new bubble.

Maples invokes Howard Marks's distinction between predicting the future and observing what everyone believes today. If investors all expect a terrible outcome, such as in commercial real estate at that moment, the consensus may already be excessive even if the underlying concern is real. Higher entry prices leave fewer paths to an attractive return; that is his framework, not a forecast that any particular asset will recover.

Asked to rate a venture AI bubble, Paley calls pricing extreme, around eight or nine out of ten. He questions whether very young companies valued at 50 to 100 times revenue have historically succeeded often. He believes excessive funding can impair a company even when its founders are talented and the cash looks like an advantage.

Paley's explanation is less about lavish spending than a limited capacity to make good decisions. More people allow a company to pursue too many experiments and scale activities that only partly work. Maples agrees, arguing that teams often hire as though they have product-market fit before finding it; a lean team can discard weak ideas and let customers pull a focused product out of it.

Paley adds a nuance: product-market fit can be a spectrum, and some successful businesses combine an adequate product with excellent go-to-market execution. The danger is accelerating a mediocre motion to justify a high valuation. More money can produce diminishing returns, yet a board may keep spending instead of acknowledging the need to slow down and fix the business.

Stebbings asks whether a company can have product-market fit without a viable business model. Maples uses a stricter definition: what can a company uniquely do that people are desperate for? In his view, funding too early may encourage teams to sell to merely interested customers instead of finding the customers who truly value the advantage.

Maples recommends preserving a large financing in a lockbox until the team has learned what works. Growth should mean accumulating increasingly attractive customers, not accepting unhelpful features or transient customers simply to hit a spreadsheet target. Otherwise, the company may build a leaky bucket and weaken its economics.

Paley says founders sometimes hear the opposite from other board members: go faster. He asks whether that means learning faster or burning faster, and says capital alone has no insight. The panel warns against treating a higher burn rate as a substitute for understanding the customer's problem.

They discuss blitzscaling as a tool for a specific situation rather than a universal rule. Paley describes Uber's exceptionally strong demand and network effects, while Maples recalls Lyft and Okta facing rivals able to satisfy the market quickly. In such cases scaling fast may be necessary; applying the same playbook before demand is clear can be destructive. Both have investment interests in the examples they discuss.

Stebbings asks how much market dominance matters when competitors have very different outcomes. Paley says early market sizing can create false precision: many of his best investments appeared constrained on paper, while obviously huge markets are often crowded. He looks for founders with deeply considered customer insights and a plausible market, rather than pretending to know its ultimate size.

Paley recalls that Trade Desk founder Jeff Green wanted the option to sell for a few hundred million dollars and avoided building a capital structure that would foreclose that choice. He also notes that a company like Pinterest could be described as serving all of commerce or simply as a digital collection site. His firm was not an investor in Pinterest; the example illustrates how easy it is to talk oneself out of an unfamiliar market.

Maples returns to entry price with the Lyft example. He says an investment of $750,000 at a $5.5 million post-money valuation bought 13.6%; at a $30 million post-money valuation, buying the same stake would require over $4 million. A fund must then be much larger or make fewer bets, changing both its return hurdle and its exposure to outliers.

Lemkin says that for his later-seed approach, higher prices mean a substantial share of investments must be winners, requiring greater evidence before the investment. Maples agrees that price and risk belong together: an expensive round with little de-risking can be survivable as a single bet but dangerous as a whole portfolio strategy.

Paley offers a hypothetical around Lyft's predecessor Zimride: a large, expensive initial round and a conventional spending schedule could have made a later pivot harder to finance. He argues that high capitalization can reduce the odds of success as well as investor returns, because later investors may be unwilling to recapitalize a very early company.

Stebbings suggests Series A could be attractive in the 2023 market because companies with revenue may be more de-risked without the same seed-stage price inflation. Lemkin challenges rigid rules, noting that venture returns still depend on outliers. A reasonable price by itself does not make an ordinary outcome extraordinary.

Maples replies that great seed outliers often start at reasonable prices precisely because they are unconventional. He says that was not necessarily a triumph of negotiation: the founders and investors saw potential others did not, and the price reflected a market for an unproven idea. He urges investors not to accept fashionable pricing as a rule of the game.

Stebbings acknowledges that some of his own portfolio has lower ownership at higher prices. Paley says his firm does not organize its strategy around a fixed ownership target, which he believes can misalign investors with founders and impede collaboration. It still considers check size and price, and will sometimes pay more for a founder it strongly wants to back.

Paley says price is one element of a broader decision, evaluated in the market context. If the next round might not exceed today's valuation, the risk and reward may be unattractive for both investor and founder. He expects to regret missing some expensive companies, but does not think that justifies paying any price whenever he likes an opportunity.

The panel considers why founders accept risky prices. A founder has to believe in the company even amid uncertainty, so a downside scenario may be difficult to internalize. Paley says some founders nevertheless choose a lower price to work with an investor they trust; Maples sees that as optimizing for the relationship, not necessarily as a calculation that the next round will fail.

Lemkin estimates that a great founder might accept a term sheet around 20% lower to work with a preferred investor, but doubts much larger differences are common. The group also debates talent: high valuations and well-known backers can signal quality to first-time executives, while experienced candidates may become more skeptical of an oversized capital stack.

Paley advises a senior executive considering a highly valued startup to ask for financials and compare the underlying business with the financing story. Lemkin says he has met candidates who did not know the relevant metrics. The suggestion is to understand the gap between the company's achievements and the price employees are being asked to believe in.

Stebbings asks what will happen to unicorns whose revenue does not support their last private valuation. He gives an example from his own fund of a company with roughly $500,000 in annual recurring revenue and a $750 million pre-money price. The speakers treat such cases as a risk to work through, not a guarantee that every company will fail.

Paley relays an LP's observation that reported venture valuations had begun to fall substantially in year-end financial statements. He offers his own interpretation about audit cycles and valuation rules rather than presenting it as a verified explanation. Private marks can lag changing conditions, so book values need not equal realizable cash.

Paley outlines three possible paths for heavily funded companies: cut burn and build toward a real exit, find a buyer at a price far below the previous valuation, or run out of money. He believes recapitalizing a tiny-revenue company at a billion-dollar prior valuation is especially difficult. These are scenarios, not determinations about individual companies.

Lemkin worries that founders burdened by unreachable expectations may quietly disengage. Paley argues that, in some cases, a founder who no longer believes in the path can discuss returning remaining cash. He distinguishes a founder-led discussion involving all stakeholders from an investor demanding its money back as if company cash still belonged solely to that investor.

Stebbings suggests that returning part of the cash can preserve a founder's time and reputation with backers. Others say large funds may be disengaged from positions that no longer look like potential outliers. Lemkin describes an unnamed promising company whose large early investor repeatedly declined invitations to attend a board meeting; this is his account, not an independently verified allegation.

Paley argues that investors should remember that startups affect employees and founders, even when the expected outcome is not fund-defining. Many successful founders built less successful companies first. Treating people well is both a responsibility and, he suggests, consistent with taking a long view of the industry.

Stebbings returns to LPs asking whether the 2021 vintage deserves a mulligan. Maples rejects a blanket exemption: he believes managers should account for their decisions and explain their strategy and mistakes to LPs. In his view, rapidly deploying a fund during a frothy period is itself a judgment that matters to the long-term relationship.

Paley says LPs must decide whether they want momentum investors or managers pursuing a longer-term strategy. He sees 2021 as a good time to sell and a difficult time to buy carefully, recalling back-to-back pitches and pressure to decide quickly during the pandemic. He warns that momentum can work during a boom and reverse sharply when the cycle changes.

Lemkin emphasizes that selecting exceptional managers is difficult for LPs too. He suggests some may need to treat a weak vintage as exceptional to manage their own relationships and funding pressures, especially when distributions have been poor. This is his interpretation of the incentives, not a rule for evaluating every fund.

On IPOs, Paley questions treating a listing as the goal itself. It can provide investors liquidity after a lockup, but often does not give founders and management immediate, meaningful liquidity. He says a company should go public when its leaders want to build for a long period and are ready for the obligations; selling can be preferable when that commitment is absent.

Paley criticizes the SPAC boom as a route that took some unready companies public, while acknowledging exceptions. He argues that a strong, prepared company may be able to list in many market conditions and that the first-day price should matter less than the ability to create value over time. His view is about company readiness, not a promise of access to public markets.

The panel tests the argument with a hypothetical Stripe listing below a previous private valuation. Paley says a past investor's entry price does not determine today's value, and public liquidity could still be useful if the business is ready. The proposed prices are a thought experiment from the 2023 discussion, not an announcement or an appraisal of Stripe.

Maples describes the unusually favorable selling windows of 1999–2000 and 2020–21, when valuations could detach from fundamentals. He says his fund chose to monetize positions in Lyft and Okta during the period of strong enthusiasm. Timing matters enormously to realized returns, though he does not claim investors can reliably predict every market turn.

The group notes that companies listed at inflated prices can leave employees facing painful declines. Lemkin then worries that lower public software multiples could cascade back through private venture valuations. He points to B2B companies with $200 million or more in revenue and argues that six- or seven-times revenue may make some later-stage venture economics difficult.

Paley counters that there is no single correct multiple. Valuation is a range affected by interest rates, growth and profitability, and different companies deserve different comparisons. He urges founders to concentrate on intrinsic value: the key question is whether their burn creates value or destroys it, rather than whether they can time a favorable market.

The speakers propose bets rather than settled conclusions. Paley expects an unusually wide gap between paper fund values and eventual distributions. Lemkin, while concerned about public multiples, predicts a strong acceleration in hiring among public cloud leaders in the following year; the others decline to present that macro view as something they can confidently forecast.

Maples likens seed investing to spotting an unusual bird: each company arrives individually, and the job is to notice a unique insight. Paley agrees that building a company can take eight to fifteen years, through many market cycles. He says the long rise before the 2022 downturn made it easy to forget how much conditions can change.

Paley says illiquidity has sometimes helped his firm's returns by preventing an early sale while a company continued to grow. He describes holding Trade Desk shares as a large shareholder and board member, with responsibilities to the company. He says the fund owned about 12.5% at the IPO and later took some liquidity; those interests shape his perspective.

Asked about LP reactions, Paley recalls advice that any decision to take secondary liquidity can look wrong in hindsight: either one sold too much or too little. He says LPs may reasonably prefer cash or stock, so managers must exercise judgment. He reports that Trade Desk burned about $7 million before going public, as an example from his experience.

Asked which holding he should have sold, Paley says his relationships with founders come first and resists naming one. Maples says his more salient mistake was selling some Twitter stock when the company was valued a little over $1 billion. He calls it a failure of imagination and says strong product-market fit is rarer and more valuable than he appreciated at the time.

Lemkin says he has preferred holding positions when he believes in the founder and wants to maximize the eventual multiple rather than optimize short-term IRR. Stebbings says his own anchor LPs advised him to retain a potentially material secondary position. Their preferences reflect their circumstances and investment interests; holding is a bet with downside as well as upside.

Stebbings thanks the panel for a lively discussion. The guests close by saying they enjoyed the conversation and should spend more time together.