This piece argues venture capital can't rely on gut feelings—it needs a repeatable framework. Jerry Neumann thinks too much money has flooded VC, pushing new funds to chase short-term traction. He sees brands and value chains (a company's unique internal activities) as new moats, citing Google beating Yahoo via a different value chain. Key holdings: The Trade Desk (his best investment, struggled for 3 years then exploded), Uber (he missed it, thinking it was just taxi optimization when it actually created new demand), and Bank Simple (an online bank that changed its value chain by eliminating branches).
Jerry Neumann, in his appearance on the Invest Like the Best podcast, explored the fundamental difficulties of venture capital. His core argument is that early-stage investing cannot rely on gut instinct but must instead adopt structured thinking. He points out that commonly cited advantages such as
Jerry Neumann is one of the most structurally minded investors in early-stage investing, and his blog reactionwheel.net is essential reading on the topic. The core thesis of this episode is: Venture capital cannot rely on gut instinct; instead, it must build a repeatable analytical framework. At the same time, traditional advantages such as network effects have been over-exploited, and brands and value chains are the new opportunities VCs should focus on.
The most impactful judgment of the entire episode: Jerry Neumann argues that making investment decisions based on gut instinct is not only largely ineffective, but more critically, it prevents you from learning from mistakes — "If you make decisions based on gut, you have no rationale. You don't know why you made that decision, so you can't improve."
Historical Context: A decade ago, Neumann used a web crawler to count active tech VCs, finding only about 100–120. Today, "800 new funds have been launched in the past five or six years," with the vast majority of GPs doing something else five years ago.
Mechanism Breakdown: New funds are small ($5M–$15M) and must quickly demonstrate performance to LPs. This leads GPs to favor companies that can show traction in the short term but have limited long-term potential—for example, in ad tech, many firms can easily grow from $0 to $5M in revenue but then plateau. Neumann cites his most successful investment, The Trade Desk: three years in, "there was almost no traction, and both the CEO and investors were anxious," but it later exploded. "After three years, you have no idea how it will turn out, but they (new GPs) have to know."
Data Chain: Neumann notes that when multiple VCs bid, "the one who wins the term sheet is, by definition, the one who overpaid." This distorts the market.
Implication: This short-term orientation systematically selects for companies that "may plateau faster," rather than those capable of building lasting competitive advantages.
Mechanism Breakdown: Traditional finance defines risk as variance/beta, but VC invests in "things that have never been done before"—if someone else has already done it, it may not be a good investment. Without historical data, probability and statistics cannot be applied. Neumann points out that "the risk VCs consider is usually how much you don't know about what is going to happen."
Key Analogy: He criticizes the approach of "reducing risk before investing." Taking food delivery as an example, investors say "I know the market is large" to reduce market risk, but the problem is: "If the market is obvious and has no barriers to entry, too many competitors will flood in, and the entire sector will be a net loss for investors." He cites the disk drive and office supply store cases from the 1980s—when Staples and Office Depot were funded by VCs, there were 12–15 competitors simultaneously, 13 of which went bankrupt.
Implication: VCs should embrace uncertainty rather than trying to eliminate it. "You embrace enough of the unknown, the unknowable, and the ambiguous, and fewer people will do it." This gives startups a "grace period"—until the market is successfully proven to exist, real competition only begins then.
Network Effects: Neumann admits he could be wrong — when he said this five years ago, Slack emerged. But he insists: "The problem space for network effects is finite; people have been actively searching for 20 years, and the opportunity space is drying up." Investing in network effect projects now is "increasingly targeting smaller, more niche markets."
IP Protection: In the tech sector, patents are easily circumvented. He cites the MP3 patent as an example: Fraunhofer Institute holds several core patents and earned hundreds of millions of dollars, but Apple made far more from MP3 than that. The reason is that Apple could say, "We use open-source Ogg Vorbis." "In tech, patents have little leverage."
Brand: Cases like Dollar Shave Club and Casper show that brand is a powerful barrier to entry. However, the problem is that brand building is extremely capital-intensive — "If you need to spend $100M to build a brand, as an angel investor I end up with only 2 basis points, and no matter how big the exit, it's not interesting enough."
Value Chain: This is the direction Neumann values most. Citing Michael Porter's theory: "Your value chain must differ from competitors and be hard for them to change — copying a product is easy, but changing internal activities is difficult." He uses Google vs Yahoo as an example: Yahoo's value chain revolved around "keeping users on the site," while Google's revolved around "getting users to leave and find things" — Yahoo could not change this system of activities. His investment in Bank Simple (an online bank) follows similar logic: traditional banks treat customers as suppliers of funds, while Simple treats them as real customers, changing the value chain by eliminating branches.
Inference: Neumann believes that the "grace period" (the window with no competition) typically lasts only a few years, after which a true moat must be built. Brand and value chain are the most promising directions to explore today.
Framework Breakdown: Schumpeter defines economic profit as "excess value beyond the cost of capital." In a competitive market, excess profits are eroded by competition, forming a curve that declines over time. The area under this curve represents the total excess value created by the entrepreneur. Two approaches: efficiency innovation (lower cost) and value innovation (higher value).
Data Chain: Efficiency innovation has an upper limit — "you can improve efficiency to 100%, but it is usually far below that." Value innovation, on the other hand, "can theoretically create infinite new value."
Key Case — Uber: Neumann admits he missed Uber because he misjudged it as efficiency innovation — "the U.S. taxi market was only a few hundred million dollars, with limited room for efficiency gains." In reality, Uber was value innovation: it created entirely new supply (a large pool of drivers), thereby unlocking suppressed demand (hailing a car within 5 minutes). "The U.S. taxi market is now at least an order of magnitude larger than it was a decade ago."
Inference: Many current SaaS companies fall under efficiency innovation — "making a certain business process 25% better" — valuable but limited in scale. Neumann focuses more on value innovation, i.e., creating entirely new markets or significantly enhancing user value.
Mechanism Breakdown: Intuition = pattern matching. Pattern matching requires two conditions: 1) the environmental signal has validity relative to past signals; 2) enough repetitions have occurred to form a pattern. VC investing satisfies neither — "How many VC investments have you made? Enough to form a pattern?" Even for someone like Neumann, who "neurotically examines every VC investment in history," the data remains anecdotal. Moreover, signals are constantly changing — "The signal I got from Sila is completely different from the one I got from another company 10 years ago."
Data Support: Citing Jeff Smart's doctoral dissertation — in 89 cases of VCs selecting founders, VCs who judged founders by intuition performed the worst, while the best performers were those who asked, "Has he done this before? Was he successful?"
Key Argument: The biggest problem with intuitive investing is not that it might be wrong, but that it cannot be improved. "If you make decisions by intuition, you have no reasons. You don't know why you made that decision, so you can't improve." Neumann cites his own miss on Uber as an example — he analyzed his mistake afterward and incorporated it into his process. Intuitive decision-makers cannot conduct such post-mortems.
Implication: Neumann believes people like intuition because it is easy and feels like magic — "People want magic." But VC investing should be like solving a puzzle: "Every investment is different; you have to solve that puzzle." He recommends a structured approach to evaluating founders: whether they understand the industry, whether they are excited, how they respond to challenges — rather than relying on "gut feeling."
| Position | Guest Stance | Key Data |
|---|---|---|
| The Trade Desk | Bullish (one of the most successful investments) | Almost no traction after three years of investment, then exploded; market cap ~$8B; company raised only ~$6M |
| Uber | Missed (acknowledged mistake) | The US taxi market was originally only a few hundred million dollars; Uber expanded it by at least an order of magnitude |
| Bonsai.ai | Bullish (investment case) | Has 15 AI PhDs, forming a "besieged resource" moat |
| Bank Simple | Invested (value chain innovation case) | Online bank, changed the value chain by eliminating branches; technology development was harder than expected |
| Unsupervised | Bullish (investment case) | Uses topological data analysis (TDA), fully unsupervised; 300-500 companies already in alpha testing; client KPIs improved by 20-30% |
| Sila | Bullish (investment case) | Cryptocurrency-based payment API; founder was a co-founder of Simple; solves the atomic operation problem of the banking system |
| Edmit | Bullish (investment case) | University tuition database; the average actual payment at US universities is 50% of the listed price; except for the top 1% of universities, all are competing for students |
| Positive reference (value chain case) | The value chain difference with Yahoo made it impossible for Yahoo to replicate Google's success | |
| Yahoo | Negative reference (value chain case) | Portal strategy: keep users on the site, leading to a decline in search quality |
| Slack | Neutral reference (network effect counterexample) | Neumann admitted that when he said network effects were exhausted five years ago, Slack emerged |
| Casper / Dollar Shave Club | Neutral reference (brand case) | Brand building was successful but capital-intensive |
| Palantir | Neutral reference (comparison with Unsupervised) | Requires setting up a dedicated department to use, deeply embedded in processes |
1. "The person who wins the term sheet, by definition, is the one who overpaid." (Jerry Neumann) — When multiple VCs compete for a deal, the winner inevitably pays a price higher than all others; this is the core mechanism of market distortion.
2. "Three years later, you have no idea what the outcome will be, but they (new GPs) have to know." (Jerry Neumann) — Small funds, pressured by LPs to chase short-term traction, systematically screen out companies with limited long-term potential.
3. "If the market is obvious and there are no barriers to entry, the entire sector will be a net loss for investors." (Jerry Neumann) — Using food delivery and 1980s disk drives as examples, a clear market opportunity leads to excessive competition.
4. "Copying a product is easy; changing internal activities is hard." (Jerry Neumann) — The essence of value chain innovation: making it impossible for competitors to respond by altering their internal activities, as doing so would disrupt their existing profit models.
5. "Uber is not efficiency innovation, but value innovation — it created entirely new supply and unlocked pent-up demand." (Jerry Neumann) — He admits misjudging Uber as efficiency innovation (only looking at the taxi market size), when in fact Uber expanded the market by at least an order of magnitude.
6. "The biggest problem with gut-feel investing is not that it might be wrong, but that you cannot improve — because you don't know why you made that decision." (Jerry Neumann) — The core advantage of structured analysis is that it allows for review and iteration, which gut-feel decision-makers cannot achieve.
7. "VCs who judge founders by gut feel perform the worst; the best performers ask, 'Has he done this before? Did he succeed?'" (Jerry Neumann) — Citing Jeff Smart's doctoral thesis, an empirical conclusion from 89 cases.
8. "A Harvard education is no better than other schools — the key is whether the individual wants to learn and works hard to learn." (Jerry Neumann) — Citing a comparative study of students near the admission thresholds of Harvard and Stuyvesant High School, concluding that the school itself does not change the person; the person is the decisive factor.