Mike Maples on Seed Investing, Founder Fit and When to Sell

Floodgate investor Mike Maples joins Harry Stebbings to discuss seed-fund economics, founder evaluation and when to sell. Both are venture investors, and Maples has financial interests in several companies he uses as examples.

Asked whether a seed fund under $100 million can work, Maples says a much smaller fund can invest at angel-sized check levels, around $100,000 or less. Its strategy and relationships must match that scale; a fund investing small checks competes differently from one leading rounds.

He says fund size is strategy because a power-law distribution makes the largest winners disproportionately important. As an illustration, he applies an 80/20 relationship recursively to suggest the top 4% might account for 64% of outcomes. This is a simplifying model, not a universal statistical law for every fund.

Maples says Floodgate's first funds benefited from timing and overlooked opportunities. Later, sitting on 22 boards left him overstretched and less able to see new possibilities. After disappointing subsequent funds, the firm discussed mistakes with its limited partners and changed how it managed follow-ons.

Floodgate assigned Iris Choi responsibility for follow-on investments and their returns. Stebbings questions separating the decision from the investor closest to a founder; Maples says Choi knows the portfolio and founders, while an independent mandate prevents a partner from pressuring her to defend a struggling investment.

Maples's principles are to be paid for the risk taken and to play offense with capital. A pro rata right is an option to invest in an existing company that others do not have. He sees value in exercising it for exceptional companies, though a fund could choose to reserve nothing.

Floodgate settled on roughly 70% of capital for initial checks and 30% for reserves. Choi is accountable for the latter and has sometimes bought more than pro rata, including in Applied Intuition. The allocation is Floodgate's choice rather than a general prescription.

Maples gives substantial weight to the judgment of strong later-stage investors. If nobody wants to fund a portfolio company, he cautions seed investors against assuming their private knowledge makes everyone else wrong; aggressive interest from a top firm is a positive signal, but not a guarantee.

He lists Floodgate winners followed by prominent firms, then notes a counterexample where Sequoia followed an investment that did not work. He likens some follow-on investing to selecting from the subset that strong firms also want, rather than blindly copying their decisions.

For his own seed strategy, Maples targets approximately 5% of first checks returning 100 times cash invested and 10–15% returning 20 times. He says achieving this can support a 10-times fund. These are targets and a portfolio model, not promised returns or a general probability for all investors.

The firm tracks initial-check picking skill separately from how much follow-on money goes into its best companies. Maples says the latter can be ranked by current value, while that value remains distinct from cash returned until an exit occurs.

Rather than forecasting a single exit value, he asks what must be true for the initial check to return 100 times. Since most investments will not land among the strongest, he considers whether the rare successful case could be big enough and whether entry price and dilution leave room for that return.

He uses Applied Intuition's $40 million post-money entry valuation as an example of a stretch price that required a very large outcome. Stebbings observes that dilution may mean the company needs to be worth billions. Maples agrees that entry price matters; later private valuations do not themselves establish a realized return.

Maples argues that active seed investors must find market inefficiencies rather than simply buy fully priced opportunities. He thinks it has become more competitive, but still sees the work as finding something others miss or gaining access to a special opportunity.

A check must also be large enough to matter to the fund. He says he would generally go no lower than about $500,000, while acknowledging a missed opportunity in ElevenLabs where a smaller check might have worked. The discussion uses a private valuation, not a cash exit.

Even an unusually expensive inception round can make sense, he says, if the founder and potential outcome justify it. He cites Applied Intuition founder Qasar Younis as exceptional. Each investment also consumes one of a fund's limited opportunities to place an initial bet.

Experience can provide a temperament advantage. Maples says he and partner Ann Miura-Ko were willing to wait in 2021 rather than invest in companies that did not meet their conditions, even as colleagues worried about the absence of new deals.

He calls this understanding one's circle of competence: know where the firm has an advantage, then vary the pace as prices and opportunities change. When others were fearful in 2009, the firm saw attractive projects; when prices rose in 2021, he says he personally made only one investment.

Miura-Ko's Lyft investment illustrates the approach. Maples recalls a $750,000 check at a roughly $5.5 million post-money valuation and estimates a very large multiple. Later in the interview they discuss actual sales and the fact that a paper multiple only becomes cash when shares are sold.

The rise of AI does not remove the need to find a price and opportunity that work, he argues. Floodgate looks for specific conditions rather than buying the whole market. He mentions enterprise-oriented investments including Applied Intuition and Cicero as examples in which the firm has an interest.

Maples has sometimes negotiated different terms from a headline round using convertible instruments. He is prepared to walk away if the price does not work. He says common stock can make sense if taking more risk materially increases ownership, and uncapped notes only in unusual high-conviction situations.

On exits, Maples recalls deciding in 2015 that Floodgate should sell some Lyft shares. The stake was deeply in the money but behind a substantial preference stack, and competition was intense. Miura-Ko sold a meaningful portion in the private market, enough, he says, to return the second fund.

He argues that seed funds can benefit from exit-price inefficiency as well as attractive entry prices. A private round may value a company as if it will execute perfectly for years. Because a small early stake can matter greatly to a seed fund, selling part can change fund economics even when it would be less significant to a large investor.

Maples suggests discussing potential secondary sales with founders well before a financing, including whether new investors would improve the cap table. He calls a sale large enough to affect a fund like an IPO an 'initial liquidity event.' His approach depends on cooperation, not selling secretly.

Stebbings raises the cost of selling too early: a great company may double again, and Bessemer's early Shopify sale is an example of upside forgone. Maples agrees that a sale may be regretted, but thinks selling part can still be the right risk-adjusted decision while retaining enough stock for further gains.

He says Floodgate also sold Lyft shares around $75 at the public-market stage, higher than the earlier private sale. That later price shows the private sale was not the absolute peak; his argument is that it reduced risk while retaining exposure.

Maples tracks companies that could return 100 times the original investment, counting both exited and unexited cases separately. He reports Twitter at over 300 times, Lyft around 205 times, and Twitch close to 94 times; Applied Intuition was approaching 100 times on the initial check by his account. The unexited valuation is not a realized return.

Floodgate had sold some Applied Intuition shares, he says, in cooperation with the founder. He favors planning sales case by case with clear criteria established before excitement takes over, rather than using a fixed sale schedule for every company.

Looking back, Maples says his biggest mistakes were often failures of imagination, including passing on Airbnb. He studies missed opportunities without treating the exercise as self-punishment: the question is what could have been known at the time and whether today's framework would produce a different decision.

His team reconstructs early investment cases for major winners, looking at potential multiples, time and dilution. It tests three signals: a distinctive insight, an external inflection and founder–future fit. Retrospective success cannot substitute for the evidence available at the original decision.

Zoom illustrates why the signals can appear uneven. Maples says its initial consumer-oriented concept was wrong, but founder Eric Yuan had spent a decade working on video conferencing at Webex. That deep exposure gave him what Maples calls founder–future fit, even though the original product vision changed.

The phrase builds on William Gibson's idea that the future is unevenly distributed. Maples looks for founders already living in a particular future, noticing what is missing and obsessively working on it. They may discover needs that outsiders cannot yet see.

An inflection is a change outside the startup that makes a new product possible; he cites GPS-enabled smartphones as an enabling condition for Lyft. The other signals are a non-consensus insight about the future and a founder especially well suited to making that future real.

For Okta, he describes Todd McKinnon's experience at Salesforce: early cloud customers were adopting more applications and would need identity management. McKinnon knew those customers and had credibility with them. Maples believes that proximity helped him identify the product and attract early believers.

Founder–future fit does not require a previous company. Maples cites Marc Andreessen working in a university supercomputer lab and building a browser while using networks and tools that many others would later encounter. His point is that direct exposure to an emerging future can matter more than conventional business credentials.

Maples says his own weakness is excessive optimism about whether people can execute. He will invest in founders he does not personally like: in his view, a breakthrough company challenges the present, and a disagreeable founder may be prepared for resistance.

When he loses faith in a founder, he tries to 'detach with love': stop repeatedly asserting disagreement, leave the relationship respectful and remain available if he can help. He does not describe this as abandoning every obligation to the company.

Stebbings asks whether the messy middle holds the most opportunity. Maples says his early focus is product-market fit and claims the companies he has seen attain durable fit have been successful. Stebbings challenges him with products that had millions of users and then faded; Maples considers those temporary spikes rather than durable fit, illustrating their differing definitions.

Maples thinks an investor can help by removing distractions and asking what bottleneck stands between the company and product-market fit, although he doubts an investor can create fit for a founder. He describes founding as a demanding vocation rather than an easy route to status.

Stebbings worries that very large rounds can push companies to spend on too many initiatives. Maples agrees that raising and hiring ahead of genuine product-market fit can weaken focus and prevent a company from learning which customers and features matter.

He predicts that many highly valued companies financed before fit will not clear their preference stacks, even if they have years of runway. This is his forecast, not a demonstrated outcome for any specific company. He notes that investors can have incentives to keep reporting progress while waiting.

Maples tells a counterexample involving CoTweet. After an investment with Steve Anderson, the company nearly ran out of money, then was acquired for ExactTarget stock. ExactTarget later went public and Salesforce acquired it; Maples says the sequence ultimately returned about 23 times the investment.

For 2024 company of the year, he selects SpaceX and points to launches and satellite connectivity. He speculates it could become an extremely valuable space platform. Stebbings raises the concentration of power around Elon Musk; Maples treats that as a separate question from his selection.

He names 20VC fund of the year, congratulating Stebbings on a $400 million raise, and picks Musk as founder of the year. He also encourages Stebbings to recognize the accomplishment while carrying responsibility for the money entrusted to him. The exchange is between professional peers.

Maples credits Charlie Munger and Howard Marks with influencing his investment thinking. He connects Marks's second-level thinking to startup investing: finding a non-consensus insight and being right about a substantially different future.

Looking toward 2025, he sees a possible future in which Bitcoin becomes more valuable than gold and supports a wider financial ecosystem. Asked for a year-end price, he guesses $130,000. This was a speculative prediction made in the interview, not an investment recommendation or known outcome.

Asked about DOGE, Maples defines success as a lasting improvement in government accountability, not simply rapid cuts. He supports setting goals and measuring results at federal and local levels; these are political and policy judgments rather than evidence that a particular initiative will achieve them.

He thinks too much money has entered venture and argues that exits are cyclical. He describes long periods in which a substantial share of profits can occur in a short liquidity window, and predicts pressure on large multistage funds if the 2020–2022 exit environment does not return soon.

Maples defines a seed round as enough money and time to test a non-consensus insight and remove the biggest initial risk. He argues that raising a large round simply because it produces a customary dilution can encourage unfocused spending and delay learning.

Stebbings counters that more money can give founders time to iterate. Maples replies that constraints can help founders discover customers and the real economics of their business. They disagree about how much early funding improves the chances of finding product-market fit.

In the final questions Maples says he values what he sees as philosophical contributions of Christianity, including unconditional love. His historical claims are his interpretation. He distinguishes loving someone and wanting their good from accepting harmful behavior without boundaries.

From his father, an early Microsoft employee, Maples learned to do his best rather than imitate someone else's strengths. He uses comparative advantage to argue that people should find the distinctive contribution they can make and honor their limited time.

Asked what he would do differently as a parent, he says he would spend more time outdoors and encourage sports. His father was intellectually focused on computers, although they did spend time together outside. Stebbings thanks him for the conversation.