The Twenty Minute VCTom Hulme: Lessons from a 24x Angel Track Record, 275x on Robinhood & Making Billions on Uber |E1150
CHAPTERS
- 0:00 – 1:04
VC emotional reality and the three investor archetypes
Tom frames venture as a daily emotional rollercoaster and introduces a simple taxonomy of investors. He emphasizes that founders should avoid the most dangerous type: those who think they add value but actually interfere.
- •VC feels like “founder on antidepressants”: highs and lows but dampened
- •Three investor types: smart/value-add, passive/non-interfering, and meddling pseudo-smart
- •The biggest risk is interference and misalignment, not passivity
- •A strong portfolio creates constant context switching and emotional volatility
- 1:04 – 3:07
Childhood, bullying, and building empathy + direct feedback muscle
Tom explains how being bullied shaped grit and empathy, and how that later informed how he interacts with founders. They discuss the tension between wanting to be liked and giving honest feedback that helps people grow.
- •Bullying built perseverance and sensitivity to unseen struggles
- •Empathy as a management/investing advantage
- •Learning to give direct feedback despite conflict-avoidance instincts
- •Best founders are receptive to feedback, especially from the market
- 3:07 – 4:47
How Tom started angel investing: experimenting with value-add vs passivity
After selling a company, Tom approached angel investing as an experiment: would he enjoy it, and would he be any good? He found he loved being on the founder ‘emergency list’ and helping with strategy/design, but notes venture’s feedback loop is painfully long.
- •Angel investing began post-exit as a deliberate learning experiment
- •Goal #1: identify investor style (value-add vs passive) and enjoyment
- •Goal #2: determine skill—hard because outcomes take years
- •“999/911 list” as a signal you’re meaningfully supporting founders
- 4:47 – 11:21
Smart vs passive capital cycles, and why deal structure can quietly hurt companies
They debate how much of the market is truly “smart smart” and how ZIRP flooded VC with passive money. Tom argues today’s risk isn’t just ‘bad investors’ but harmful structures—notes, preferences, and ratchets—driven by incentives like TVPI marking.
- •ZIRP increased passive capital; tighter markets reward fundamentals again
- •Convertible notes used to avoid repricing and protect reported TVPI
- •LP/GP incentive chains propagate optimistic marks
- •More structure: heavier liquidation prefs and IPO ratchets
- •Founder advice: optimize for options and total terms, not just headline price
- 11:21 – 13:40
Performance, regime change, and why angels often shouldn’t follow on
Tom shares strong early angel results but argues they’re hard to interpret because the market regime changed and forecasting is unreliable. He explains why, as an angel, he concluded follow-ons can worsen outcomes—later rounds become a different competitive game.
- •Pre-2015 angel batch: ~4.5x DPI and ~24–25x TVPI across ~27 companies
- •Regime shift: early valuations then vs now make comparisons misleading
- •He would have stack-ranked his own portfolio incorrectly
- •Angel strategy: often avoid follow-ons; pro rata can be expensive and misread signals
- •Reserves models can overestimate our ability to predict winners
- 13:40 – 17:19
What winners had in common: fundamentals, long compounding, and avoiding ‘hot’ rounds
Tom contrasts decade-long compounders with momentum-driven stories that looked great briefly but collapsed. They argue ‘hot’ seed/Series A rounds can be negatively correlated with long-term success by distracting founders and inflating expectations and burn.
- •Best outcomes came from difficult, fundamental businesses compounding for years
- •Example: GoCardless growing steadily into a major ARR business
- •Momentum/heat plays can create paper wins with fragile foundations (Fab/Jawbone examples)
- •Hot early rounds can distort founder behavior and raise premature scaling risk
- •Long-term compounding beats short-term narrative-driven value
- 17:19 – 20:51
Zeros and bad angel decisions: social proof traps and founder diligence frameworks
Tom unpacks mistakes behind failed investments: chasing heat, outsourcing diligence, and not spending enough time understanding founder thinking. He shares recurring questions he uses to evaluate founders, focusing on unfair advantage, timing, and paranoia.
- •Common failure mode: momentum investing and relying on others’ diligence
- •Social validity: big-fund small checks can create misleading ‘A+’ signal
- •He probes product to understand how founders think, not the product itself
- •Key questions: unfair advantage, ‘why now?’, and what could go wrong
- •Strong founders can list risks and ask for help; lack of paranoia is a red flag
- 20:51 – 25:26
Ideas are cheap; execution, timing, and ‘cockroach mode’ determine survival
They debate whether a great idea without ‘secret sauce’ is enough; Tom insists execution dominates because good ideas get copied. He discusses timing risk—being early is like being wrong unless you can survive long enough—and highlights bootstrapped patience before adding fuel.
- •Investing requires belief in execution advantage, not novelty of idea
- •Market timing risk: ‘too early’ equals ‘wrong’ without endurance
- •VR as a cautionary example of waiting for adoption curves
- •Bootstrapped endurance can outperform overfunded early scaling
- •Minimum learning cycle is deals + time, not a dollar threshold; start with small checks
- 25:26 – 26:37
Founders as angels vs founders running side funds: focus and fiduciary risk
Tom supports founders investing their own money because they can be genuinely value-add specialists. He’s concerned about founders running side funds because it’s effectively building another business with real fiduciary responsibility on top of an already brutal job.
- •Founder-angels (own money) can be ‘smart smart’ with empathy and domain value
- •Side funds add a second company worth of workload and distraction
- •Raising others’ money increases responsibility and potential misalignment
- •Signal risk: what does a side fund imply about commitment to the core company?
- 26:37 – 31:23
The four S’s of great venture: sourcing, selecting, supporting, and selling
Tom describes VC as three core jobs—sourcing, selecting, supporting—and adds a fourth: selling. He explains how senior investors risk losing sourcing relevance and shares his own strengths (enthusiastic sourcing/network) and weakness (over-investing emotionally in support).
- •Three S’s: sourcing, selecting, supporting; plus selling as an underrated core skill
- •Career drift: senior investors often shift toward supporting and risk losing sourcing edge
- •Tom’s edge: enthusiasm enables broad sourcing and network-driven referrals
- •Hard part: supporting without being pulled into every founder crisis
- •Great board members act as shock absorbers for founders
- 31:23 – 38:45
Networks, pre-mortems, and sizing winners: optimism with disciplined downside thinking
Tom argues ‘serendipity favors the connected’ and acknowledges VC is partly a game of access. He explains how optimistic cultures need structured pessimism via pre-mortems, and why investors routinely underestimate winners because they fail to imagine multiple S-curves.
- •Network effects drive opportunity flow; luck often reflects connectivity
- •Access matters, though venture can still be a partial meritocracy once inside
- •Pre-mortems (Kahneman): create permission to articulate how decisions fail
- •Investors underestimate winner magnitude due to linear S-curve thinking
- •Scenario planning framed as option value: data, distribution, and unlocked paths
- 38:45 – 44:12
Liquidity drought, fund-return dogma, and ‘never sell winners’ vs taking some off the table
They discuss ecosystem liquidity constraints with IPO/M&A windows constrained and PE filling gaps. Tom challenges rigid ‘must return the whole fund’ thinking, emphasizes strategy consistency, and balances GV’s long-hold orientation with pragmatic advice to take partial liquidity when available.
- •Reduced exits threaten the ‘multiplier effect’ that recycles founders/operators into new startups
- •Rocket-ship alumni are advantaged founders (ambition + operating cadence)
- •Tom disputes that every deal must be a full fund return; alternative models can work
- •GV optimizes for long-term absolute returns more than near-term IRR
- •Regret-minimization: selling 10–20% is often rational to avoid future regret
- 44:12 – 51:45
GenAI investing reality check: commoditization, foundation model economics, and cloud utility endgame
Tom argues the genAI market is bifurcated: frenzy for AI while everything else is hard. He believes foundation models resemble rapidly depreciating power stations with limited defensibility, likely becoming utility-like offerings controlled by hyperscalers, making fundamentals-based returns difficult.
- •GenAI looks like ZIRP-era behavior again; valuations can be disconnected from durable value
- •Foundation model training = massive capex with fast obsolescence and marginal differentiation
- •Meta’s scale + open source (LLaMA) accelerates commoditization pressure
- •End state thesis: models become a cloud utility; hyperscalers monetize compute/distribution
- •He would struggle to invest in OpenAI at ~$90B absent stronger defensibility (memory, agency, etc.)
- 51:45 – 1:00:00
Where durable value may exist: application layer moats and commercialization capability
Tom lays out what he wants in genAI applications: proprietary data/distribution or enterprise-grade productization that benefits as models improve. He also argues many AI-native teams lack commercialization experience, and estimates a high failure rate—especially in foundation models.
- •Application-layer durability requires data, distribution, or full-stack enterprise readiness
- •Example lens: Synthesia building security + GTM + end-to-end product in a new category
- •Altman test: be ‘happy’ if the model improves 100x (not devastated)
- •Under-asked question: can the team commercialize and iterate fast, not just research?
- •Zero-rate guesses: ~90% foundation models fail; ~70% application layer fail; incumbents capture much of value
- 1:00:00 – 1:09:52
Founder craft: naive outsider vs insider, clock speed, pricing for feedback, and culture debt
From a founder perspective, Tom says both naive outsiders and experienced insiders can win if they compensate for gaps with humility and hiring. He stresses execution clock speed, charging early to get real feedback, and managing culture debt—especially post-COVID and in hybrid environments.
- •Naive founders can win via speed and recruiting domain depth (Lemonade example)
- •Insiders can win via deep buyer understanding (CurrencyCloud/Visa example)
- •Startups are ‘unanswered questions’; great founders choose order and answer fast
- •‘Free kills feedback’: charging validates value and yields higher-quality signal
- •Culture debt is harder than tech debt; hybrid often worse than fully remote or fully in-person
- 1:09:52 – 1:17:50
Quick-fire: boards, Elon/Neuralink, robotics, defense tech, parenting, and backing founders twice
In rapid questions, Tom shares his favorite board experience (Nothing), memorable pitches (including Neuralink with Elon), and shifting views on robotics. He also discusses why defense collaboration matters, how parenting reframed his view of ‘nature vs nurture’—and why that maps to not trying to ‘change’ founders.
- •Best board: Nothing—global perspective and customer-first execution
- •Memorable pitch: Neuralink; impressed by Elon’s first-principles technical depth
- •Changed mind: robotics becoming more generalizable via AI + cheaper components
- •Contrarian belief: working with the military will matter more amid geopolitical risk
- •Parenting lesson: founders’ core traits are hard to change; best to support and position them well
