Uncapped with Jack AltmanInvesting with Conviction | Sarah Guo, Founder of Conviction | Ep. 5
CHAPTERS
- 0:01 – 2:27
Defining what a venture firm really is: money, people, beliefs, and advantage
Jack and Sarah unpack why a VC firm feels “wispy” compared to operating a company. Sarah frames venture as a bundle: commoditized money plus the squishier components—people, beliefs, brand, and tangible advantage founders can feel.
- •Founders are the customer; VC’s value proposition should be analyzed through that lens
- •Capital is mostly a commodity; differentiation comes from people, beliefs, and help
- •Brand and individual reputations feel fragile but can still compound over time
- •Enduring firms suggest some elements of venture can become durable institutions
- 2:27 – 3:59
Why top VC firms endure: ethos, tribal knowledge, brand, and network
They explore what compounds at firms like Sequoia and Greylock across cycles and generations. Sarah argues durability comes from an inherited ethos and learned pattern-recognition, amplified by brand and networks that are hard to replicate quickly.
- •Venture isn’t purely zero-sum; new firms can be built without “attacking” incumbents
- •Enduring firms preserve an internal ethos across investor generations
- •Tribal knowledge includes both useful pattern recognition and biases
- •Brand and founder networks are slow to reproduce and provide staying power
- •Large platforms also benefit from available capital, though capital is abundant generally
- 3:59 – 8:07
How founders choose VCs: brand halo vs individual partners (and it changes with experience)
Jack asks whether different founders optimize for different VC attributes, including brand. Sarah contrasts first-time founders needing trust and signaling with repeat founders who often choose specific individuals regardless of firm logo.
- •Founders are not homogenous in what they value from investors
- •Some founders optimize for prestige/halo to recruit and sell (brand as leverage)
- •Repeat founders often pick specific partners rather than firms
- •Stage and founder experience change how much brand vs help matters
- •Conviction aims to attract a particular founder set rather than everyone
- 8:07 – 10:44
Building Conviction’s brand intentionally: solving information asymmetry with “partner marketing”
Sarah describes a formative loss to Andreessen and how it shaped her thinking about VC perception. She explains Conviction’s deliberate brand strategy: make the network and value visible at scale, much like a startup marketing funnel.
- •A founder’s blunt comparison: firms look similar; some feel bigger, better at PR, and pay more
- •Early truth: “nobody cares”—you must create reasons to care
- •Brand as a tool to reduce information asymmetry about network and capabilities
- •Tactics resemble partner marketing: show up with high-distribution partners (e.g., major AI/tech players)
- •Brand work can also directly help portfolio companies, especially at early stages
- 10:44 – 13:08
What Conviction stands for: being AI-native and willing to take risks with opinions
They move from tactics to philosophy: Conviction’s brand is rooted in having a point of view and taking risk publicly. Sarah highlights publishing LP letters and making predictions as a feature, not a bug—founders can opt into that worldview.
- •Core pillar: willingness to have an opinion and take risk on it
- •AI-native understanding is a current “proof point,” not necessarily the permanent topic
- •The firm’s name signals enduring ethos: conviction can travel to future technology waves
- •Publishing beliefs invites scrutiny but provides grounding in dynamic markets
- •Brand should communicate worldview so founders can self-select alignment
- 13:08 – 16:33
Do brands drive returns, or do returns drive brands? What “success” means to founders
Jack probes the causality between brand and performance. Sarah argues brand should ultimately be grounded in quality and success of companies, but founders may not (and shouldn’t) optimize for a VC’s fund multiples—those are proxies for taste and learning.
- •Best brand foundation: quality of companies, not just marketing presence
- •Founders often lack visibility into true “how early” or ownership (primary vs late secondary)
- •A VC’s returns aren’t perfectly aligned with what founders care about (e.g., dilution)
- •Returns can proxy for taste, pattern recognition, and association with quality
- •Sarah believes learning compounds more from success than from failure
- 16:33 – 20:52
The 2025 venture landscape: fund growth, distorted incentives, and the challenge for small firms
They map today’s market dynamics: mega-funds, the post-ZIRP hangover, and why raising capital can be easier than generating top-tier returns. Sarah explains structural incentives that push firms to grow and why that may degrade performance without shrinking the asset class.
- •Capital markets are deep; fundraising can outpace actual return generation
- •Fee economics can reward larger funds even with weaker multiples
- •Growth is an unnatural default for venture but organizationally self-reinforcing (career ladders)
- •Backwards-looking performance and under-allocation to venture keep capital flowing in
- •The industry may resemble private equity: hard to outperform, but capital doesn’t necessarily leave
- 20:52 – 22:43
Can small funds still win when big platforms subsidize early-stage? Conviction’s bet
Jack asks whether large firms can rationally “overinvest” at seed to fill pipelines, hurting small-fund economics. Sarah acknowledges the risk but argues the early market is not homogeneous—small firms can still compete by being distinct and needing only a few outsized wins.
- •Big multi-stage firms can subsidize seed strategy with later-stage economics
- •Potential thesis risk: early-stage may become a funnel with compressed alpha
- •Conviction’s constraint (small fund, small team) forces selectivity, not volume
- •Founders choose different experiences: bespoke partnership vs large platform
- •You don’t need many wins to generate a great outcome on a $200M fund
- 22:43 – 27:25
Where alpha is (and isn’t): taste, willingness to choose early, and different winning strategies
Jack asks where the best risk/reward pockets are today. Sarah downplays “global optimization” and emphasizes playing to one’s strengths—early-stage taste—while noting that some investors win by owning communities or being deliberately controversial and early.
- •Sarah focuses on being excellent at early stage rather than chasing stage-level arbitrage
- •Mainstream narratives can flip fast (e.g., “no app-layer money” to “only app-layer money”)
- •Winning strategies vary: brand-building, community ownership, or contrarian theses
- •Examples of alpha: controversial early bets that later become obvious
- •Early-only funds must “choose” before markets consolidate and entry prices explode
- 27:25 – 31:11
Lessons from Greylock: linear market thinking, transitions, and picking the right people
Sarah reflects on learning investing at Greylock and what she carries forward. She highlights the power of understanding existing markets, tracking macro transitions, and knowing talent networks—often rooted in incumbents—while still leaving room for new-market creation.
- •Greylock as a training ground; admiration for Asheem’s capital preservation + upside
- •Value of linear thinking: clear logic tied to market structure and transitions
- •Traditional pattern: identify major spend categories → understand tech shift → find the right people
- •Networks in incumbent markets can be a repeatable edge
- •VC remains personal: company outcomes and investing styles reflect people and taste
- 31:11 – 34:01
Market-driven vs founder-driven in AI: new categories beyond traditional software spend
They contrast traditional market-led investing with AI’s ability to open non-software markets. Sarah cites examples where AI enables products that replace services or create entirely new capabilities, making some “old” market maps less predictive.
- •AI shift creates important markets outside classic enterprise software categories
- •Examples: legal workflows, customer support work, video creation, healthcare provider tech
- •Prior skepticism about some sectors (e.g., digital health) shifts when value becomes undeniable
- •AI’s value often shows up as “technology does way more,” not just cheaper software
- •Traditional market focus can miss emergent categories that don’t look like software yet
- 34:01 – 36:28
From inputs to outputs in AI: adoption signals, capital efficiency, and what matters for Conviction
Jack references Satya Nadella’s reframing toward measurable economic output. Sarah agrees it’s rational for hyperscalers to demand revenue justification, while noting Conviction’s job is different: find venture-scale returns with a smaller pool of capital, anchored by real business traction.
- •Satya’s framing: massive model CapEx must be justified by economic output and shareholder returns
- •Hypothesis: owning the “fast takeoff” lab may not guarantee economic value capture
- •Conviction’s bar is different: sufficient value creation to return a top venture multiple on $200M
- •Preferred signals: revenue, durability, and capital-efficient growth (even EBITDA makes an appearance)
- •Not trying to be a CapEx-heavy foundation model fund; strategy matches fund size
- 36:28 – 42:42
More $10B+ companies and the “AI eats services” mechanism—plus the pricing of agents
They debate whether today’s valuations imply a future with many more giant outcomes than history suggests. Sarah argues AI expands willingness-to-pay by moving beyond SaaS budgets into labor/services budgets, while acknowledging future pricing pressure and commoditization for non-differentiated offerings.
- •AI era may genuinely create more very large companies than prior waves
- •SaaS lesson: adoption grew, but budgets and “software appetite” eventually capped out
- •AI may shift spend from software budgets to labor/services budgets (new willingness-to-pay)
- •Agents price against labor today; long-term pricing likely pressured toward compute/intelligence costs
- •Differentiated products (Figma-like uniqueness) may retain value-based pricing longer
- 42:42 – 46:29
Agency as the scarce human advantage—and what to teach kids in an AI world
They close on the idea that intelligence is becoming abundant while agency remains distinctly human. Sarah ties this to founder selection (force of will, point of view, predictive judgment) and to parenting: focus on behaviors and reasoning skills that make people adaptable amid shifting tools and institutions.
- •Beyond intelligence, Sarah prioritizes force of will and the ability to take—and be right about—a point of view
- •Agency as shaping the world through action feels harder to automate than raw intelligence
- •For kids: emphasize frustration tolerance, concentration, and the ability to upskill
- •Structured reasoning (decomposition/debugging; STEM-style thinking) still matters even with AI tools
- •Remain flexible about future education paths as institutions and learning modes evolve
