Uncapped with Jack AltmanSeed Investing at Scale | David Tisch, Managing Partner at BoxGroup | Ep. 10
CHAPTERS
- 0:00 – 0:26
BoxGroup’s “favorite investor” ethos and what VC really is at different stages
David frames BoxGroup’s goal as being founders’ “favorite investor” by being human, honest, and not overpromising. He also distinguishes seed, Series A/B, and later-stage investing—arguing later-stage is closer to a finance job while seed is inherently messy and art-driven.
- •“Favorite investor” = talk like humans, don’t mislead, underpromise/overdeliver
- •Venture contains multiple distinct businesses (seed vs A/B vs B+)
- •Later-stage investing: capital allocation into proven traction; more finance-oriented
- •Seed investing: embraces messiness and uncertainty as the core feature
- •BoxGroup intends to stay seed-focused rather than evolving upmarket
- 0:26 – 6:33
Scaling a collaborative seed fund without becoming a lead/ownership-driven firm
Jack probes why conventional VC wisdom says you must move later-stage and concentrate to scale, and David explains why BoxGroup didn’t. The model is designed to prioritize access to the best companies over ownership targets, avoiding a “pick me or die” dynamic.
- •Many firms scale AUM by moving later-stage, writing bigger checks, and leading rounds
- •BoxGroup stays in seed because the mission is to invest in the best companies
- •Ownership requirements create conflicts with that mission and force competitive behavior
- •Collaborative participation reduces friction: “Pick us and them,” not “pick us”
- •Consistency matters because founders are on 10–15+ year timelines
- 6:33 – 8:23
Power laws, portfolio size, and why “shots on goal” isn’t the real strategy
They discuss power-law outcomes and the intuition that more bets increases odds of outliers. David clarifies BoxGroup isn’t maximizing volume for its own sake; it’s saying yes whenever a company clears a quality bar, acknowledging that most decisions will still be wrong at seed.
- •Most startups fail; seed investors must accept being wrong frequently
- •Failure represents real human cost—founder dreams, emotions, and restarts
- •Concentrated models assume an unrealistic ability to be consistently right
- •BoxGroup invests in many companies because they see many great companies
- •Pace of investing should flex with opportunity quality, not a quota
- 8:23 – 11:22
Why stack-ranking seed companies is mostly ego—and how VC narratives get rewritten
Jack describes the impulse to rank companies precisely, and David argues early stack-ranking is not realistically possible. They criticize how the industry highlights winners, hides losers, and rebrands itself around new themes, creating misleading narratives about investor foresight.
- •Early-stage outcomes are too uncertain to reliably rank beyond broad tiers
- •Claiming you “knew the winner immediately” is often ego and hindsight bias
- •Most firms build baskets, then market only the successes
- •Failures and once-high-conviction misses get “swept under the rug”
- •Industry attention resets quickly (e.g., shifting to the newest theme like AI)
- 11:22 – 17:20
The real seed bar: founder quality, leadership potential, and “does it matter if it works?”
David defines the seed decision as fundamentally person-driven rather than market/traction-driven. He explains the “butterflies” moment of meeting exceptional founders and the core evaluation: can they lead and recruit at scale, and would the outcome be important if successful?
- •Seed is primarily about the person, not spreadsheets, markets, or traction
- •Occasionally you back founders even with skepticism about the initial idea—if they’re exceptional
- •Founder evaluation: IQ+EQ package and ability to lead/recruit at massive scale
- •Second filter: importance—if they’re right, does it matter (consumer vs enterprise vs regulated domains)
- •Timelines are extreme (7–15+ years); founders and funds must align to that patience
- 17:20 – 22:30
Hiring for taste: why it’s selected, not taught—and why BoxGroup avoids IC politics
Jack asks whether investing “taste” is teachable; David says it’s hired for and shaped by life-wide preferences and experiences. He describes BoxGroup’s trust-based structure where any investor can say yes, rejecting committees that create politics and deal-negativity.
- •Taste is broader than venture reps; it’s formed through life choices and interests
- •BoxGroup hires for taste rather than trying to train it from scratch
- •Team stability builds trust; many seed firms churn and become inconsistent
- •No voting/IC: any team member can make an investment decision
- •Consensus models can introduce politics, incentives, and founders-facing opacity
- 22:30 – 28:59
The art of being collaborative: compounding relationships and owning fundraising help
David explains collaboration as an extension of a founder-first principle: figure out what each founder needs and help them get it. BoxGroup focuses on being excellent at helping companies raise follow-on capital, building durable relationships across the funding ecosystem as stages compress.
- •Collaboration starts with “help the founder,” not “be the hero”
- •Fundraising is the universal need across a portfolio; BoxGroup optimizes for it
- •Follow-on capital has moved earlier; seed now overlaps with multi-stage investors
- •Being non-competitive enables authentic, compounding relationships over years
- •Consistency and discipline over 15 years is harder than constantly reinventing the model
- 28:59 – 37:31
“VC help” is overrated: value is mostly access, brand, and not distracting founders
They unpack what VCs can realistically contribute, emphasizing that operational “company building” is rare. David argues the best value at scale is network access (capital, customers, hires), while overinvolved investors can damage focus and confidence—especially for first-time founders.
- •VCs often over-attach to successes (“my companies”) and overclaim impact
- •Most “help” is introductions/access; true operational help is rare
- •Platform promises can distract founders; second/third-time founders avoid noise
- •Wrong advice is damaging because it shakes founder conviction and focus
- •BoxGroup aims to respond to what founders ask for—especially fundraising support
- 37:31 – 48:12
Signaling, brand effects, and why some VCs pass for unspoken reasons
David argues “signaling” is often misunderstood: it matters mainly when companies struggle, not at the top of the market where preemption happens. He also explains that VCs usually pass because the deal didn’t clear the bar—often team quality—though they rarely say that directly.
- •Signaling is most real in the bottom half: when insiders won’t support a struggling company
- •Top-of-market rounds get preempted; prospective leads won’t wait for insider intent
- •Brand is uniquely powerful in venture and affects fundraising, recruiting, and customer trust
- •VC “pass reasons” are often polite soundbites; root causes are usually team/idea excitement
- •“Too early” is often code for “I want more proof,” but the market can move overnight
- 48:12 – 54:54
NYC vs Bay Area: ecosystem depth, where scale happens, and learning from USV
David separates where BoxGroup is based (New York) from where it must invest (often the Bay Area). He explains ecosystem “depth” as the limiting factor for building and scaling to thousands of employees, praises USV as a defining NY firm, and argues geography matters more for life than for investment scope.
- •A NY-based firm can’t ignore Bay Area value creation; limiting to NYC “cuts off the head”
- •NYC attracts ambitious builders but can be “hard mode” for certain company types
- •AI era recenters talent density in San Francisco; depth is the advantage
- •Scaling talent needs (10k+ employees) are only feasible in a few ecosystems
- •USV as an underappreciated benchmark: thesis-driven, low-FOMO, consistently excellent
- 54:54 – 59:55
North stars: team trust, aiming for irrationally big outcomes, and respecting founder-scale wins
In closing, David shares the principles he does hold tightly: build a team with deep trust and autonomy, and only back things that can become truly huge. He also highlights a dual reality—venture math needs massive outliers, but “smaller” exits can still be profoundly life-changing for founders.
- •Primary north star: a partner team with implicit trust, happiness, and individual excellence
- •Autonomy enables decisions beyond what one leader could personally see or make
- •Venture only works if some investments can become extremely large outcomes
- •Compounding at the end matters most (public-market growth can dwarf IPO valuation)
- •Respect the emotional meaning of any founder win, even if it doesn’t drive fund math
