David Frankel on Seed Investing, Liquidity and Backing Founders

Founder Collective managing partner David Frankel joins Harry Stebbings to discuss seed investing, reserves, liquidity and competing with incumbents. Both are venture investors; Frankel's firm owns or has owned stakes in many companies he discusses, and their opinions reflect those interests.

Frankel believes smaller seed funds can still find non-consensus opportunities despite larger rounds. He points to founders from less conventional backgrounds, those returning after failure and neglected markets. His portfolio company Smalls, a cat-food business, is an example; he reports about $50 million in ARR when it struggled to raise, a company figure rather than independently verified revenue.

Seed pricing still follows a market, he says, but he will often decline highly priced AI rounds that cannot fit his fund economics. He recalls making an unusual $100,000 exception for a founder he liked. A small check can make sense, he says, only when expectations about the investor's time and involvement are clear.

The size of a potential return relative to the fund affects how much time an investor can rationally devote. Frankel thinks founders should understand that a large fund may reserve its deepest engagement for very large outcomes, and that a prestigious initial investment does not guarantee follow-on support.

He recalls passing on Pinterest because of a conflict with an existing portfolio company. Although the outcome makes the decision painful in hindsight, he regards loyalty and conflict rules as obligations rather than merely financial calculations. He says Founder Collective remains cautious about backing direct competitors.

Early signals can mislead. Frankel says he was once pessimistic about Olo before it developed, and a consumer company's sudden success can make an investor overestimate a founder's abilities. He emphasizes patience and tolerance for a long, uncertain journey.

Founder Collective initially had no reserves. Frankel recalls partner Eric Paley breaking that rule to support The Trade Desk when it needed money, with a follow-on of roughly $500,000 to $1 million. The later success is an exception, not evidence that every rescue investment will work.

The firm subsequently adopted a reserve strategy, roughly one-to-one in his description, but has found it difficult to allocate perfectly. Initial rules limiting follow-ons by valuation became outdated as market prices changed. He says forecasts of future portfolio needs are inherently hard.

At the start of COVID, the team feared funding would disappear and tried to conserve money for its companies; the market then became awash with capital. Frankel treats reserve management as one of venture's most difficult tasks because the environment can change against the assumptions behind a fund plan.

Unlike investors who avoid hard feedback, he says he is direct when a company is running out of cash. He recalls telling Running Tide's founder that burn was too high; the business later shut down. This is his account of a difficult decision, not a full assessment of the company's options or technology.

He thinks founders may find it hard to cut spending because of loyalty to the team, investor expectations and the same determination that made them start the company. A prospective large customer can also keep hope alive. His lesson is to manage monthly burn while there is still time to act.

He has sometimes found founders returning after failure especially compelling: they may be hungry, experienced and able to bring a strong team back together. He describes Motorway founder Tom Leathes as an example, and says investors should explicitly invite founders they respect to return with their next idea.

Frankel criticizes pro rata rights as a free option for investors that can complicate a founder's next financing. If an existing investor waits for someone else to set terms, a potential new lead may hesitate; when a new investor wants meaningful ownership, existing rights can also increase dilution. The effect depends on the round and actual agreements.

Stebbings notes that an early commitment can itself be used to seek better terms elsewhere. Frankel sees that as a useful early signal about the relationship, though founders also have legitimate reasons to compare offers. He contrasts it with investors making unconditional early commitments to show strong support.

He supports a basic preferred-stock right to return investors' capital before common shareholders, but warns that many later rounds can leave earlier holders behind a large preference stack. He says tougher financing conditions can bring stricter terms. Outcomes depend on the exact seniority, participation and documents.

Public-market software multiples had fallen from the 2021 highs, and Frankel says some founders had difficulty accepting that the market had changed despite company progress. His advice is to watch cash burn rather than assume the next round can be raised at the previous valuation.

He worries about distributions to paid-in capital, or DPI, for funds from 2018 onward. He says Founder Collective's 2018 fund had not yet distributed cash at the time of the conversation, unlike earlier funds. That is a fund-specific status, not proof that distributions across the asset class are permanently dead.

He describes a period of cheap money and rapidly escalating valuations followed by a difficult exit market. A private-equity letter of intent for one portfolio company was still only a proposed transaction; he expects some liquidity to come through PE acquisitions, but neither that deal nor a wider recovery was guaranteed.

Frankel hopes one or more strong IPOs could reopen the market, while Stebbings believes some vintages will face permanent capital losses rather than mere delays. Both are discussing uncertain future outcomes and the difficulty LPs face waiting for distributions.

A small seed fund need not rely solely on a single enormous outcome, Frankel argues. Several medium-sized exits can collectively return a fund if ownership and proceeds are sufficient. His examples are illustrative and depend on actual stake, dilution, fees and distribution timing.

He cites Uber, The Trade Desk, Coupang and PillPack among important outcomes, and describes SeatGeek and Airtable as potential future fund returners. He says PillPack returned the second fund. Potential returns from remaining holdings are not realized distributions.

A founder can assess investor incentives by comparing the check and potential outcome with fund size. Joining the board can signal commitment, although Stebbings notes that some large-fund partners do not show up consistently. Frankel says some do and cautions against treating fund category as a complete substitute for checking the individual.

Limited partners respond differently to illiquidity: institutions may maintain some venture allocation, while family offices may pull back further. Frankel says his small fund can accommodate an LP leaving if its minimum check is too large for the allocation. These are his observations, not a comprehensive survey.

Secondary liquidity is often available for the best-known private companies but difficult for smaller holdings, he says. At IPO, Founder Collective tends to distribute large positions as stock so LPs can choose whether to hold, while it may sell smaller positions and distribute cash. Each choice leaves exposure to a different market outcome.

He regrets some sales of Trade Desk and Uber shares because those businesses later grew. He acknowledges that hindsight makes the decision look easier than it was. His preference to hold exceptional companies longer is an opinion, not a guarantee of future performance.

Frankel calls certain entrenched incumbents 'lethargic economic extractors causing harm,' or LEECHes. He argues that lobbying, legal action and public relations can make disruption difficult. This is his adversarial characterization as an investor in challengers, not an independent finding about every incumbent.

He recounts a SeatGeek sales visit at a Boston venue and says he later gave the account to the Department of Justice. He alleges the venue manager described pressure tied to access to Live Nation concerts. The interview presents his recollection and allegation; it does not independently establish the incident or a legal violation.

He also discusses record-label litigation against Suno, a Founder Collective portfolio company. He presents Suno's view that training on material from the open internet is analogous to learning from music. Copyright liability is disputed and the interview does not establish that the training is lawful or resolve the case.

Frankel thinks investors can help challengers prepare legal, lobbying and communications resources, but says the first requirement is a product customers actually value. Stebbings asks whether incumbents always survive; Frankel expects they may remain powerful while their market shares change.

He still wants evidence of product-market fit from an AI team before committing heavily, though he recognizes scaling and distribution can require substantial capital. Asked about a hypothetical large investment in OpenAI, he speculates it could return twice the money or more; this is a prediction, not a disclosed result.

Most very high-priced AI seed rounds fall outside his fund's normal economics. He recalls passing on a first round at a $250 million valuation because dilution and fund size made the required return difficult. Missing a generational winner is possible, but he does not believe a seed strategy can depend on identifying that exception every time.

He expects AI's near-term financial returns to disappoint some investors relative to infrastructure spending, while predicting profound longer-term change. He compares adoption with the slow progress of self-driving cars. Both the timing and scale of the AI outcome remain uncertain.

Frankel recalls an LP asking how his partner knew Uber would succeed. His partner answered that he did not know, and had similarly high hopes for surrounding investments. Frankel sees this as a reason for humility: even successful investors cannot identify every eventual outlier in advance.

Founder Collective asks whether a deal has a plausible path to a tenfold return. Frankel says a small fund can be returned by several such successes and need not demand that every company become a giant. The target is a screening heuristic, not a forecast for each investment or fund.

On losses, he says a mistake is investing in a compelling idea when he does not really believe in or enjoy working with the founder. By contrast, a strong team can still fail because timing or market conditions were wrong. His preference for personal chemistry is a judgment, not an objective measure of founder ability.

A reduced allocation can be acceptable when it brings in useful collaborators, he says; context matters. He does not equate a hot deal with a good one. When demand for a round is extreme, a secondary sale may be possible, although he also regrets selling some exceptional holdings too early.

Frankel supports modest founder secondary sales when they relieve personal financial pressure and preserve focus. He says investors benefit too, and criticizes growth investors who encouraged very large sales in a boom and later blamed founders. His suggested percentage is a rule of thumb, not a universal standard.

He regards excessive capital as a central mistake of the prior boom: large rounds could encourage international expansion before a local model worked. He recounts transactions involving Tiger as examples, while acknowledging he was caught up in the excitement. These are his recollections, not proof about any named deal's eventual outcome.

For incumbents with valuable customer data, he sees AI as an opportunity to improve and defend their products. He expects underlying tools to become cheaper and doubts every customer can build its own vertical software. Whether companies can raise prices enough to offset AI costs remains an open question.

On board work, Frankel values economic alignment because investor time is limited; for his fund he suggests a meaningful initial stake, often around 15% or more. He also says he continues helping founders he likes even when ownership is lower, showing that relationships can matter alongside the economic calculation.

He prefers to dilute alongside founders rather than automatically maintain ownership through expensive rounds. He sees board value in timely advice and connections, especially when a company faces unfamiliar legal or operational problems. His comparison with larger funds is his perspective as a seed investor.

He describes receiving expert guidance through a portfolio bankruptcy involving alleged fraud, and says he returned secondary-sale proceeds. The company is not identified in the interview; his account should not be generalized into findings about other companies.

Frankel calls a seed investor a matchmaker: helping founders reach the right partner at the next fund can be central to the job. He argues that large funds may have different incentives because they can invest in later rounds themselves. That is his incentive analysis, not proof that any named firm never makes introductions.

Stebbings adds that investors can introduce people to repay favors rather than because they are the best fit. Frankel says context matters: when financing is difficult, a willing investor may still be valuable. Founders need to understand the relationship and their own alternatives.

Market sizing matters, but early estimates can miss future expansion. Frankel recalls MediaRadar growing beyond the newspaper and magazine market initially assessed by investors. The lesson, in his view, is to test whether a founder can find adjacent markets without treating the eventual opportunity as certain at the start.

He has changed his mind about private equity, citing portfolio companies where financial discipline and operational help benefited shareholders. He expects PE to provide some liquidity, while recognizing transactions still have to be negotiated and completed.

He names Jake Saper as a valued board colleague, recommends patience and says he wishes he had understood how long the journey can be, including after an IPO. He says Running Tide was his biggest loss and that the firm lost its investment when the product did not become sufficiently commercial.

The hardest decisions often feel close rather than obvious, he says. Moving from South Africa to build Founder Collective involved weighing both opportunity and what he would leave behind. He tries to combine analysis with the emotional response he has to a founder.

Asked what matters most, Frankel names love, friendship and loyalty. He recommends David McCullough's The Path Between the Seas as a history of the Panama Canal and reflects on the ambition, financial stakes and human cost of large projects. Stebbings thanks him for the conversation.