Uncapped with Jack AltmanInvesting with Conviction | Sarah Guo, Founder of Conviction | Ep. 5
CHAPTERS
Setting the stage: why venture firms feel “wispy” compared to startups
Jack frames the core question: a company has a clear product-to-revenue loop, but a VC firm seems less tangible. Sarah sets up how she thinks about venture from the founder’s perspective and why the job still matters despite the squishiness.
- •VC is harder to define than an operating company
- •Founders are the “customers” of a VC firm
- •Venture’s differentiation can feel fragile and point-in-time
- •Despite this, enduring venture institutions can be built
What a VC firm is at its core: money, people/beliefs, and advantage
Sarah breaks venture down into a bundled value proposition. Money is commoditized, while people, beliefs, brand, and practical advantage are the real differentiators—though harder to make durable.
- •Venture as a bundle: capital + associated people/beliefs + advantage/help
- •Capital is a commodity; differentiation is mostly non-monetary
- •Brand and individuals are “squishy” but can still compound
- •The best firms create enduring value beyond capital
Why the best VC firms compound over decades: ethos, tribal knowledge, brand, network
They discuss what persists at firms like Sequoia and Greylock across generations. Sarah argues durability comes from institutional ethos, learned patterns (and biases), and compounding networks and brand reputation.
- •Durability is not purely zero-sum; new firms can exist without “replacing” old ones
- •Enduring elements: ethos passed across generations
- •Tribal knowledge: what worked, plus biases shaped by tech history
- •Brand + founder network are hardest to replicate quickly
How founders choose investors: brand vs individual partners, and how needs change by stage
Sarah explains that founders aren’t homogeneous: some buy prestige/halo effects, others pick specific people. Repeat founders may care less about the firm logo and more about who they want in the trenches with them.
- •Different founders optimize for different VC attributes
- •Prestige brand can help with recruiting/customers (halo)
- •Repeat founders often choose individuals over firm brand
- •Conviction aims to attract a specific founder set, not everyone
Building Conviction’s brand intentionally: solving information asymmetry with “partner marketing”
Sarah shares an experience losing a deal to Andreessen that shaped her thinking about differentiation. She describes brand building as making founders care and demonstrating the network in scalable, accessible ways—similar to a startup marketing funnel.
- •Founders often perceive VC firms as similar aside from size/PR/check size
- •“Nobody cares” → firms must create awareness and trust
- •Demonstrate network rather than asserting it in 1:1 persuasion
- •Use distribution via partnerships/events (e.g., with major tech players)
Opinionated investing as identity: why “Conviction” is about taking risks on beliefs
Beyond being “AI native,” Sarah positions the firm’s ethos as being willing to have opinions and take risks. Publishing LP letters exemplifies this: some predictions will be wrong, but having a worldview can be grounding for founders in a dynamic era.
- •Core brand pillar: willingness to take a stance and be non-neutral
- •AI-native credibility is part understanding, part network, part execution help
- •The name “Conviction” signals opinion-taking beyond any single trend
- •Publishing beliefs creates accountability and attracts aligned founders
Do brand and returns cause each other? What “success” means to founders
They probe whether strong brand leads to strong returns or vice versa—and whether founders should care about VC fund multiples at all. Sarah argues returns are mostly a proxy for taste and quality; founders ultimately care about association, learning, and outcomes that help their company.
- •Brand should ultimately be rooted in company quality and success
- •Founders often can’t see how early/meaningful a VC’s involvement really was
- •Fund returns aren’t directly what founders should optimize for
- •High-quality hits imply taste, useful tribal knowledge, and strong association
The 2025 VC market map: big platforms, spin-outs, and distorted growth incentives
Sarah gives a candid read: raising money is easier than making money in venture. Structural incentives push firms to grow AUM like businesses, while LP behavior lags due to backward-looking performance from pre-ZIRP vintages—creating likely return compression.
- •Capital markets are deep; fundraising is easier than delivering top returns
- •AUM growth is tempting and organizationally reinforced (career paths, expansion)
- •LP demand is influenced by backward-looking best-vintage multiples
- •Industry maturation (like private equity) may persist despite weaker returns
Can small early-stage funds still win when mega-firms subsidize seed?
Jack asks if mega-firms distort early-stage economics by using seed as pipeline. Sarah acknowledges the risk but argues early-stage remains non-homogenous: small firms can win by being distinct, selective, and high-conviction without needing huge headcount or capital.
- •Potential bear case: early-stage becomes a feeder funnel for large platforms
- •Large firms can subsidize seed with resources and lower cost of capital
- •Conviction’s strategy: selective focus—only a few breakouts needed per fund
- •Founders value different experiences than working with a giant institution
Where alpha comes from: taste, controversial bets, and choosing earlier without “waiting”
Sarah downplays optimizing for “where the best risk-reward is” across stages; she focuses on what she’s good at—early-stage taste. They discuss alternative winning strategies (community ownership, contrarian picks) and why early-stage firms must choose before markets settle.
- •She doesn’t try to globally optimize stage-by-stage alpha—leans into early-stage strength
- •Main job: build the ability to win deals you want (brand + distinctiveness)
- •Alpha can come from being intentionally controversial/non-consensus
- •Early firms can’t wait for categories to resolve; must pick amid uncertainty
Lessons from Greylock: market understanding, transitions, and matching talent to shifts
Sarah credits Greylock and Asheem with teaching her core investing patterns. She highlights a powerful traditional approach: understand incumbent markets, identify technology transitions, and back the right people—often emerging from incumbents.
- •Greylock as foundational training; Asheem’s record spans preservation, doubles, home runs
- •Traditional pattern: map big existing markets, spot macro transitions, pick the right teams
- •Incumbent-market networks and category knowledge are compounding advantages
- •This isn’t “opposite” of AI investing—just adapted to faster-changing frontiers
AI creates new software markets: from enterprise categories to services and “non-software” spend
Sarah argues AI enables companies to attack markets that weren’t historically software spend categories (law, support, creative services). They contrast SaaS’s growth limits (software saturation) with AI’s potential to expand what’s worth buying—unlocking new budgets.
- •AI-native companies often sell into previously non-software markets (e.g., law, support)
- •SaaS misprediction: cross-sell saturation—organizations hit “enough is enough”
- •In areas like healthcare, prior tech often wasn’t worth buying; AI changes value delivered
- •“AI eats services” reframed as technology doing far more valuable work
From inputs to outputs: Microsoft’s pragmatism, signals of adoption, and agent pricing pressure
They discuss Satya Nadella’s reframing toward measurable economic output rather than AGI rhetoric. Sarah emphasizes revenue and capital efficiency as adoption signals, while acknowledging that agent pricing is currently benchmarked to labor and will face long-term competitive pressure.
- •Satya’s stance: focus on economic growth outputs vs speculative AGI capture
- •For Conviction, the bar is simpler: enough value to return a great fund on $200M
- •Key signal: real revenue and durable businesses; capital efficiency matters
- •Agent pricing may start pegged to labor savings but should compress with competition unless uniquely differentiated
Agency as the “last human resource” and what kids should learn in an AI world
Sarah is skeptical that intelligence alone was ever sufficient; she prioritizes force of will, point-of-view formation, and navigating uncertainty as founder traits. For children, she focuses on behaviors (focus, frustration tolerance, upskilling) and structured reasoning (decomposition/debugging), while staying flexible about future education paths.
- •Founder selection: intelligence plus force of will and the ability to hold/validate a POV
- •Agency as shaping the world—harder to automate than raw intelligence
- •Kids: emphasize concentration, frustration management, and self-upskilling
- •Structured reasoning (STEM-like decomposition/debugging) remains valuable; education pathways may change
