This interview features Neil Mehta of Greenoaks Capital, who explains how he picks future S&P 500 companies. His core idea is finding firms that create "jaw-dropping customer experience" (JDCE), like Coupang, which built 12-hour delivery in Korea and boosted customer retention from 30% to 60%+. He also looks for founders with "credible aggression"—ambition backed by a real plan. Key holdings mentioned: Coupang (added to position, customers say they can't live without it), Carvana (bought as stock crashed from $300 to $5), and TripActions (went from $100M revenue to zero; he invested $500M during COVID).
Neil Mehta left D.E. Shaw at age 27 in 2012 to found Greenoaks Capital, allocating 40% of his initial $50 million fund to the Korean e-commerce company Coupang, which ultimately generated returns of approximately $8 billion. Over 13 years, Greenoaks has invested in companies such as Figma, Wiz, Carv
Neil Mehta, founder of Greenoaks Capital, left D.E. Shaw at age 27 to create the fund, which manages nearly $15 billion in assets, focuses on just 55 core companies, and is operated by 9 investment professionals. In the interview, he systematically elaborated on his investment framework: seeking companies that can become important components of the S&P 500 in the future, with core criteria being "Jaw-Dropping Customer Experience" (JDCE) and exceptional founders. Core judgment: Mehta believes that "competition today is actually less than it was ten years ago," because most institutions optimize for coverage rather than deep insight, and through extreme focus and differentiated information depth, Greenoaks has almost no real competitors in this "narrow track" of 10–15 best founders per year.
Mehta believes that the core of investing is to find companies that have created "jaw-dropping customer experience" (JDCE), and the starting point of such an experience is breaking a "trade-off" that others thought impossible.
JDCE is the "internal totem" of Greenoaks. Mehta says, "If you get the Greenoaks people drunk, they will talk about JDCE all night." The core logic is: in the capitalist world, the vast majority of companies are just "me-too products"; true value creation comes from the very few companies that "delight humanity by the degree of differentiation."
Coupang is the birth case of JDCE. In 2013–2014, founder Bom Kim decided to shift from a marketplace model to first-party (1P) — building its own warehouses, delivery, and routing systems. At that time, Korean consumers said, "We don't need faster delivery" (2.5–4 days was enough), but Kim insisted on achieving "12-to-24-hour delivery." Mehta points out that this seemed simple but was extremely difficult: the unit economics were "completely broken" from the start, requiring the construction of new warehouse management systems, new routing software, localized delivery points, and even research into "how to leave packages at the doorstep at 6 a.m. without waking up the customer." This process took 2–4 years before the flywheel began to turn.
Data confirms: Before the 1P capability (Rocket) was launched, the average customer retention rate in the Korean e-commerce market was around 30%; Coupang's Rocket delivery customer retention rate was above 60%. More critically, there was the emotional response of customers — Mehta played a customer video produced by Coupang: "A mother said through tears, 'The diapers arrived at my door in the morning, and I no longer have to carry a big box home from the store — if you take this service away, I don't know what I'll do.'"
Mehta's summary: "This is not an NPS score of 9; this is a jaw-dropping customer experience."
Mehta believes that great founders share a recognizable "archetype," with core traits including extreme focus, a true understanding of "what it means to say 'no,'" and "credible aggression."
Focus: Take Bom Kim as an example. His schedule was "filled with just one thing from Monday to Friday." If it meant negotiating the cost of diapers in mid-2014, he would spend six hours a day on it for two consecutive weeks, "letting everything else burn." Mehta notes: "People like to say they are focused, but most don't really understand what focus means — focus means saying 'no' to everything else."
Credible aggression: Mehta distinguishes between "every founder saying they want to climb the highest mountain" and "credible aggression" — the latter is reflected in the founder having already mapped out a route, demonstrated a resource allocation plan, explained the possible trade-offs, and even tried several paths and adjusted along the way. This is not "pure ambition" but "ambition with a route."
Key lesson about "negative information": Greenoaks' biggest early mistake was missing out on SpaceX. Mehta admits: "We heard negative things about him (Elon Musk) — he fires people quickly, he goes down to the F-level, he disappears and comes back to change everything. Our mentors said, 'He's not worth investing in.' We didn't do our own homework and listened to external opinions." Mehta reflects: "Now we know that some of those traits are exactly what we look for — we like micromanagers, people who dive deep into details, and people who fire quickly."
Mehta believes the investment industry is undergoing "private equitization," but for the 10–15 best founders, this is precisely "the wrong way to operate." Greenoaks competes through extreme focus, contrarian action, and an informational advantage that goes beyond first principles.
Divergence from Industry Mainstream: Mehta observes that most institutions now operate in a "matrix" structure — divided by sector and geography, optimizing for "coverage." He calls this model "correct for over 90% of the industry," but "for the 10–15 best founders each year, it is completely wrong."
Greenoaks' Differentiation:
Carvana Deep Dive: The stock price fell from $300 to $5, and the market believed "bankruptcy was only a matter of time" (loss of $3,000 per unit in EBITDA plus $2,000 in interest = $5,000 cost per unit). Mehta's judgment: Ernie Garcia (CEO) was not inactive but was running internal A/B tests to determine the optimal cuts. When Mehta flew to Phoenix for dinner with him, Garcia almost teared up talking about "the employees' children hearing at school that their parents' company was going bankrupt," rather than his own situation. Mehta said: "In the chaos, he slowed down the pace — that convinced me he would navigate through safely."
Mehta argues that "growth is an output, not an input," but "irrational high growth is extremely beneficial for a company's long-term health, especially for software companies."
Core thesis: "There's a saying that's become popular recently: 'Too much growth destroys companies.' I believe irrational high growth is very good for companies." Mehta quotes Mario Andretti's famous line: "If everything seems under control, you're just not going fast enough."
Case 1: Wiz During the Israel-Hamas conflict in October 2023, a large proportion of the team was called up for military service. Greenoaks' instinctive reaction was to "cancel the November-December expectations and just focus on the people and the country." In the end, Wiz "turned in one of its best quarters ever." Mehta says: "Having irrational expectations is a competitive advantage."
Case 2: Coupang One year, growth fell to 18% (year-over-year). Some suggested, "Keep this pace, Amazon never exceeded 30%." Mehta persuaded Bom Kim: "One of the hardest things is convincing the entire company to become a high-growth company again." Kim later told him: "This was one of the hardest things, and I'm glad we did it."
Growth Persistence: Mehta points out that historical data shows the best tech companies not only grow fast but also have extremely high "growth endurance" — rapid growth can persist for a very long time.
Mehta presents a counterintuitive view: despite the massive surge in industry capital (from Don Valentine’s $200 million to today’s $20 billion), “competition has actually decreased for Greenoaks.”
Logic breakdown: Most institutions are “adding complexity” — segmenting by industry, geography, and stage, optimizing to “not miss a single Series A.” This causes them to lose “fidelity.” Mehta offers a key observation: “We have been doing Series B for 13 years, and in almost every round there is another company trading in the same year at a similar valuation. The best companies and the worst companies are almost identically valued at the Series B stage — almost no one can tell the difference.”
Why? Because “if you and I evaluate Coca-Cola, the difference in our EPS forecasts 10 years out is not that large. But in this industry, both companies are doing $30 million ARR, growing 100% year-over-year, one might be worth billions and the other might be dead in five years — yet they are valued almost identically at the Series B stage.”
Conclusion: Greenoaks competes through “differential insight” rather than “coverage.” “If you chase coverage, you sacrifice depth. I have never seen a company with more than a handful of truly great investors.”
| Ticker | Guest's Attitude | Key Data |
|---|---|---|
| Coupang | Bullish (continuously increasing position) | 40% of initial fund bet (~$20M); total investment less than $1B; held to date, continued buying after IPO; customer retention improved from 30%+ to 60%+ |
| Carvana | Bullish (contrarian buy) | Stock price fell from $300 to $5; per-unit EBITDA loss of $3,000+ interest of $2,000; Greenoaks bought from $50 down to $5 |
| TripActions (now Navan) | Bullish (contrarian investment during COVID) | Revenue went from $100M to zero; Greenoaks committed $500M; rose from 4th-8th in industry to Top 2 |
| Rippling | Bullish (rapid support over SVB weekend) | Agreed to $500M investment within 30 minutes; ensured all client employees received salaries on time that Monday |
| Wiz | Bullish (no explicit position action) | Team was called up during conflict yet delivered best quarter |
| SpaceX | Missed/Regret | Did not invest early due to negative external reviews, called "biggest mistake" |
| Tencent/QQ | Early observer (no explicit position) | QQ added 30 million monthly subscribers (around 2008) |
| DoorDash | Indirectly mentioned (no explicit position) | Case of Masayoshi Son (SoftBank) supporting Tony Xu during tough times |
| Figma, Stripe, Discord, Toast, Flipkart | Bullish (Greenoaks holdings) | No specific data provided |
1. Neil Mehta believes "the investment industry is becoming privatized, but this is the wrong way to operate for the 10-15 best founders each year" — Most institutions cover by matrix (industry × geography), optimizing for "not missing out" rather than "deep understanding", leading to almost no difference in valuation for excellent companies at Series B, but vastly different future value.
2. "JDCE is not a user experience score, but the customer's emotional feeling: 'If you take away this service, I don't know what to do'" — The case of a Coupang mother crying and saying "Don't take away Rocket Delivery" defines JDCE better than any NPS score. It comes from breaking the "impossible" trade-off.
3. "Attention is the dimension with the fewest competitors — because most institutions add complexity rather than reduce noise" — Mehta believes that investing is essentially a "game of reducing complexity", "you only need an IQ of 110", but you need "gut toughness" to allow things to be that simple.
4. "Growth is an output, not an input. But unreasonably high growth is extremely beneficial to a company" — Coupang returning to 30%+ growth from 18%, and Wiz achieving its best quarter during a conflict, both prove that "having unreasonable expectations is a competitive advantage".
5. "When a CEO is at the edge of life and death for the company, see if he only cares about himself" — When Carvana's stock price went from $300 to $5, Ernie Garcia, almost in tears, talked about the feeling of "employees' children hearing at school that their parents' company is going bankrupt", rather than his own situation. Mehta believes this is a key signal worth betting on.
6. "Greenoaks' 'biggest mistake' was listening to others' negative comments about Elon Musk" — Missed SpaceX because traits like "he fires people quickly and gets into details" were interpreted externally as "uninvestable". Mehta's lesson: "Never outsource your judgment."
7. "Volatility is a signal we welcome — because that's when information asymmetry is greatest" — During COVID, TripActions' revenue went to zero, and the SVB weekend crisis at Rippling — these were action moments when Greenoaks committed $500 million in 4-30 minutes. Mehta's logic: "When others panic, the endgame judgment hasn't changed."
8. "Buffett can sell 30 IQ points, we can sell 40-50 IQ points — it's not complicated, the key is discipline" — Mehta believes that the investment framework has no secrets, "just doing the same thing over and over, looking at thousands of companies, and only screening those 10-15."