David Tisch and Terrence Rohan on Seed Investing, Conviction and Founder Relationships

Terrence Rohan tells Harry Stebbings that he spent seven years leading the seed practice at Index Ventures after working at Google. He has backed companies including Figma, Notion, Hugging Face and Robinhood. His fund, Otherwise, provides capital to a network of founders who make investments, while Rohan also invests directly, generally with $250,000 checks.

David Tisch says BoxGroup, which he has run for roughly 15 years, recently closed its sixth early-stage fund at $212 million and has a separate follow-on fund. It typically writes $500,000 to $1 million checks at pre-seed and seed, is comfortable being a smaller investor in a round and does not take board seats. He calls its approach collaborative seed investing.

Asked why founders choose an investor, Tisch says the first task is meeting them early and building relationships, including with repeat founders. Most investors offer the same basic product—money for equity—and he is skeptical that a brief fundraising pitch can prove the future value of a decade-long relationship.

Stebbings asks whether small funds can compete with major multistage firms. Tisch says BoxGroup seeks to be an additional partner, not a substitute, because aggressively competing could damage its ability to work with others. He wants founders to choose the firm because they like and trust its people.

Tisch wants BoxGroup to be a founder’s favorite investor: someone the founder would gladly work with again and can approach as a friend. Rohan agrees that introductions and other value-add services can be commoditized; trust through the ups and downs of building a company matters more to both investors.

Tisch says venture is frequently reduced to tactics and statistics, even though most early-stage investments fail and that failure affects a person who had a dream. He believes an investor must be present for both difficult and successful outcomes. Both guests have a commercial interest in their respective investment models.

Rohan sees a seed market ripe for change. He points to more funds and managers, generational turnover at established firms, and founders gaining more tools, capital options and bargaining power. He also sees AI creating uncertainty about which firms will lead and how early-stage companies raise money; these are his assessments and predictions.

Tisch emphasizes the shift to remote meetings around the pandemic. In his view, founders can reach major firms and get decisions much faster than when travel and physical meetings constrained access. That has narrowed an old speed advantage of early-stage specialists and increased founder options.

Tisch argues the smallest first rounds remain highly fragmented and relationship-driven, making the seed market difficult to index or aggregate. He distinguishes sub-$2 million initial rounds from larger rounds often called seed. Stebbings challenges whether multistage firms’ large checks are becoming more common; Tisch and Rohan say such checks have existed for years.

On price, Tisch argues that rigid valuation rules can exclude a rare, exceptional company. A $10 million round at a $100 million valuation could be a mistake if the company ends at $200 million, but attractive if it becomes a $100 billion outcome. These are hypothetical outcomes, not predictions or evidence that price never matters.

Rohan says the power law makes conviction in an exceptional company rarer than a particular price. Tisch adds that valuations once considered extreme can later look cheap, while acknowledging that a portfolio’s blended entry valuation matters. Both favor judging the company and possible eventual exit over an automatic seed-stage price ceiling.

Tisch distinguishes earlier investing, which he sees as more art and judgment, from later-stage investing, where financial analysis plays a larger role. He notes exceptions and says he would not want an entire seed portfolio entered at an average $50 million valuation; conversely, a very low average could mean missing sought-after companies.

The job remains hard because investors cannot know which founders, ideas and markets will produce extraordinary outcomes years later. Tisch questions claims that seed investors can be right dramatically more often than peers, while acknowledging exceptional investors such as Roger Ehrenberg. Rohan stresses humility about luck and the limits of prediction.

Tisch says founders and employees, not investors on podcasts, build companies. He disputes the idea that investors can confidently claim they knew every outcome in advance. At the same time, he says seeing promising opportunities and developing access are skills; Stebbings points out that content marketing can help with that access, which Tisch readily acknowledges.

The two investors say the market has expanded across geographies and types of business, so no one sees as much of it as when seed was a smaller ecosystem. Fast rounds make coverage harder. Tisch tries to meet founders before a financing process, sometimes before they have started a company, to build a basis for a decision.

Tisch says it is difficult to improve picking when definitive results can take five to ten years. He prepares by learning markets and meeting people, but often says yes when a founder’s insight and commitment make him excited about an ambitious idea. He treats this as judgment, not a reproducible formula.

Rohan believes group decisions can optimize for consensus and safety when seed investments often depend on an individual’s hard-to-articulate intuition. Tisch agrees: at BoxGroup, an investor can say yes without a firm-wide vote. Colleagues discuss opportunities and share knowledge, but the people closest to a deal make the decision.

Tisch says it is easy to list reasons a young company will fail, but seed investors must find reasons one might work. In a power-law portfolio, he views missing an eventual outlier as potentially more costly than making an individual unsuccessful investment. Rohan warns that formal votes and informal group pressure can extinguish fragile early ideas.

BoxGroup discourages colleagues from throwing a destructive “grenade” into a deal discussion. Tisch says excessive negativity can favor safer, less varied outcomes. He also notes that moods and external conditions affect intuition, so maintaining a consistent strategy and a clear head is part of the work.

Rohan describes two parts of Otherwise: founder investors and his own investments. He believes prominent founders can see opportunities through their networks and win allocations because other founders seek them out. He also argues that investing in people and domains they know can improve their picking, though he gives no comparative performance data.

Tisch agrees founders can have advantages in access and winning, but says Rohan’s model depends on selecting strong founder investors; not every founder is a good investor. Rohan responds that many exceptional founders naturally invest within their networks and expertise, while acknowledging exceptions.

Rohan generally avoids reserves for follow-on investments. He argues they can reduce distributions to paid-in capital (DPI): investors may be cut back in the strongest rounds, misidentify winners early, or put substantial capital into a different, difficult game. He also prefers the relationship clarity of a single initial check. These are his arguments, not demonstrated fund-return results.

Rohan says standard limited-partner expectations often favor 30 to 50 investments and a reserve strategy, making alternatives harder to explain. Tisch agrees follow-ons are difficult but says his separate fund can make good decisions at each round and aims to invest in the best opportunities, often relatively early. Rohan acknowledges a trusted early investor may earn access later.

Tisch warns that no venture playbook makes returns automatic. In his view, established firms’ brands and investors help them see and win opportunities, while BoxGroup seeks trusted relationships that can also provide access to follow-on rounds. His preferred reserve strategy is to invest before the broader market recognizes a company’s quality.

Not every successful company follows the same path, Tisch says: some work early, while others take time to find product-market fit. He describes roughly 150 companies in the seed portfolio and says each requires its own understanding rather than simple pattern recognition. He does not claim past examples make the next winner obvious.

Tisch says allocating between initial checks and reserves, and deciding when to move to the next fund, are difficult. In an illustrative $100 million fund with a 2% fee over its life, he describes $20 million of management fees and a goal of putting $100 million or more into companies through recycling and other means. Stebbings questions whether the current liquidity environment permits that; Tisch says it is hard. He regards over-reserving as the larger mistake.

Both guests say they generally do not sell secondary shares. Rohan thinks holding can help the fund’s return profile, although selling a small part of a mature position need not damage a founder relationship. He expects secondary markets to remain an increasingly useful option, especially if mergers and acquisitions contract.

Stebbings asks whether selling 20–30% of a position in a $10 billion company could return cash while retaining upside. Tisch says selling can make sense if little upside remains, but he values the compounding of an exceptional company over time. He does not categorically rule out a secondary sale; the quoted valuations and percentages are examples.

Tisch says staying relevant requires finding new founder networks and understanding new fields. Earlier founder relationships and reputation compound if the people he backed recommend the firm. Rohan expects AI and other tools to make some companies more capital-efficient, though he believes an initial financing round will remain important and later venture capital can become optional.

For founders, Tisch’s advice to investors is to make and communicate decisions quickly, avoid wasting time, and help when asked. Rohan says the strongest founders may need investors least; he tries to back exceptional people while recognizing he cannot know in advance whether each will build an exceptional company.

Rohan advises emerging managers that raising a fund moves much more slowly than closing startup investments. He says many rejections and a 12-to-18-month process can be normal, especially for a first fund. Tisch adds that a manager should have a reason and a vision for the firm, because returns can take a long time.

Asked about missed investments, Rohan says a company that looks successful today can still fail, so he is most disappointed when he has not followed his own principles. Tisch avoids naming a company, noting that earlier founders disliked being singled out and that apparently bad passes are not necessarily settled outcomes.

Tisch believes venture returns and access tend to follow a power law, with leading firms often retaining advantages although challengers can emerge. Rohan emphasizes brand: in his conversations, some founders choose a multistage firm partly because its name can help validate a company to potential hires. Both have interests in their firms’ reputations.

The guests disagree on signaling. Tisch says a multistage investor’s decision not to lead the next round is often less decisive than people claim, especially when competitive deals move quickly; a prolonged failed fundraise may create problems regardless. Rohan says the existing investor’s posture can affect deal momentum and that founders are asked about it.

Rohan says that at Index he tried to equip founders with a clear explanation of the firm’s position when they raised a new round. Tisch counters that a competing firm often cannot wait for perfect information without risking the deal. Rohan considers signaling most consequential for companies in the middle, rather than obvious successes or failures; Tisch continues to dispute the label and its causal importance.

Tisch contends that only a small number of exceptional investors can substantially change a company’s trajectory; most are primarily providers of capital and some can be harmful. Rohan agrees that at seed the creative work must come from the founder. Their critique reflects their view of investing, not a measured estimate of investor impact.

Tisch dislikes hierarchical labels such as coach or mentor for investor-founder relationships, preferring reciprocal relationships. He acknowledges that outside coaching can help leadership and culture, especially later, but says company success primarily depends on founders and employees. Stebbings cites HubSpot’s account of Sequoia helping with talent and customers; Tisch says a handful of exceptional investors can move the needle.

On misalignment, Rohan says the ownership and board-seat product a VC sells to limited partners may not match what a founder wants. Tisch worries that an intrusive investor can impose decisions without actually improving the company. Both say private, asymmetric information about competitors can be useful, while Tisch is skeptical of forwarding public articles founders are likely to see anyway.

Tisch accepts cold inbound and says BoxGroup has missed good opportunities sitting in its inbox. Rohan considers a thoughtful, personal cold email a form of founder hustle, unlike a mass-produced pitch. A deck or product can help convey an original vision, although he focuses heavily on the person.

Both say they can invest through uncapped SAFE agreements. Rohan mentions provisions for conversion if a company does not raise by a certain point; Tisch rejects rigid rules that might exclude a company he wants to back. Terms, risk and eventual outcomes depend on the individual agreement and company.

Looking toward 2034, Rohan expects more of the present market, possibly with more automated tools, but believes the core remains personal relationships and early bets on people. Tisch expects larger multistage firms to remain involved while small seed funds can still back outliers. Both predict that the messy, human process of starting companies will resist complete market efficiency; neither presents this as a certainty.