Kunal Shah on winning in India, second-order thinking, the philosophy of startups, and more

ELI5/TLDR

Kunal Shah — the philosophy major who built CRED and sold an earlier startup for ~$450 million — sits down with Lenny to explain why building in India is a different game with different rules. The core idea: a country’s quirks (low trust, no concept of an hourly wage, arranged marriages, low per-user income) aren’t trivia — they dictate what kind of products win and what kind of companies can exist. He threads everything through two lenses: a simple test for whether a product will actually take off (“Delta 4”), and second-order thinking — the habit of asking “and then what happens after that?” Along the way he uses Hindu mythology to explain CEO succession, crocodiles to explain corporate survival, and lab-grown diamonds to explain how status collapses.

The Full Story

The Delta 4 test: is your product four points better, or are you kidding yourself?

Most founders are told their product must be “10x better.” Shah finds that useless — you can’t measure it, and “you don’t know if you are 10x better or not unless you’re delusional.” His replacement is almost embarrassingly simple. Rate the old way of doing something out of 10. Rate your new way out of 10. If the gap — the delta — is four or more, you have a winner.

The classic example: rate hailing a cab the old way (maybe a 3) versus Uber (maybe a 9). Gap of six. When the delta clears four, three things happen.

Every time you see that the product efficiency Delta is greater than equal to four, three things happen. It is irreversible. Second is that you have very high tolerance for it to fail. And the third thing is that I call it the UBP — unique brag-worthy proposition.

Irreversible means you’ll never go back. High tolerance for failure means if Uber glitches once, you forgive it instead of quitting. And brag-worthy means you can’t shut up about it — which is why genuinely great products have low or zero customer-acquisition cost. Nobody discovered ChatGPT through an ad. Someone showed them.

The flip side is the warning. Buying a suit online is a “tech product” stuffed with features, yet the experience is often worse than walking into a shop — Lenny rates both a 5. Delta of zero. So it’s reversible, nobody brags, and you’ll happily never do it again. Shah’s blunt lesson for founders whose “better Excel” isn’t catching on: being a bit better isn’t enough. People are busy. If you’re not four points better, they won’t even look up.

Why so many Indians run American tech giants

Microsoft, Alphabet, Adobe, IBM, Palo Alto, Starbucks, pre-Elon Twitter — all run by India-born CEOs. Shah offers conjectures, not proof. The hunger of immigrants who came from nothing. A society that prizes mathematics and engineering so hard that “philosophy major was a write-off” — meaning the people who make it out the front door have already cleared brutal filters.

But his favourite explanation is mythological, borrowed from author Devdutt Pattanaik. Picture a 2x2: values on one axis, obedience on the other. Lord Rama sits high on both — he follows the rules, he upholds dharma. Krishna is high on values but low on obedience — the “naughty god” who breaks rules to create. The villains (Ravana, Duryodhana) are the ones low on values. The great hired CEOs, Shah argues, are masters at flipping between Rama and Krishna while never abandoning values — and crucially, they preserve the founder’s dharma rather than stamping their own ego on the place.

A lot of CEOs have this need to say “oh I’m going to change the company forever and it will be my identity.” … But maintaining Dharma — “these are the principles given to me and I’m going to sustain this and make it even bigger” — comes from humility.

Tim Cook maintains Steve Jobs’s dharma. Satya Nadella, during the OpenAI board crisis, “played Krishna” — broke the rules to fix things — then switched back to Rama to keep scaling. Lenny connects this to Jim Collins’s “Level 5 leadership,” where the leader’s ego disappears behind the institution. Shah’s deeper point: chasing personal status — needing your signature on everything — is itself the seed of creative destruction.

Long-term cultures are risk-averse cultures

Why does India fear startup failure so much? Because it’s a deeply long-term society — 95% arranged marriages, under 1% divorce, families built on values rather than attraction. “All bad behavior in humans comes from being short-term,” Shah says; cultures that survive thousands of years do so by everyone thinking long. The cost of that patience is timidity about risk.

He illustrates with a cruel little parable. At a big FMCG company, the person handed a risky zero-to-one launch tries, fails, and gets nothing — no promotion, no poaching offer. Meanwhile the colleague who babysat an established brand with natural tailwinds, doing essentially nothing, keeps rising. “We do not celebrate risk-takers as a country yet.” His hopeful counter-image is a church in Portugal where only royalty and the great explorers — Vasco da Gama and his kind — were granted the highest resting place. Give risk-takers the highest social status, and a country gets more of them.

The India playbook: easy users, hard rupees

The single most useful business insight in the conversation: in India, DAUs (daily active users) are cheap and ARPU (average revenue per user) is brutal. ARPU is basically a function of per-capita income — “you cannot make $100 per user from a country where income is $2,500 a year.” India offers the world’s cheapest data and huge smartphone penetration, so global giants harvest enormous user numbers here (Shah guesses Meta makes maybe $3–4 per Indian user per year) to flatter their growth charts. Netflix, Spotify, Amazon Prime all arrived expecting tens of millions of paying customers and didn’t get them — too much free content, too little willingness to pay for time.

This breaks the Western gospel of focus. In Silicon Valley you build one beloved thing, then expand. In India, because each user is worth so little, you must do many things. And there’s a second reason: India is a low-trust market. Weak institutions mean consumers won’t sue a coffee shop that injures them, and won’t gamble on unknown brands. Counterintuitively, low trust concentrates trust into a few mega-brands.

The lack of trust creates concentration of trust. You will see one app can do 400 things. We have a company like Tata that can do salt to car to jewelry to anything.

Hence super-apps thrive in developing markets. Brand isn’t just important in India — it’s exponentially more important. (The world’s oldest brand, he notes, may be Chyawanprash, named after the sage Chyawan — trust flowing from a person’s reputation, the way J.P. Morgan once did.)

The country that has no word for efficiency

A quietly startling observation: no Indian has ever been paid by the hour. Ask anyone — restaurant staff, senior executive — their income per hour, and they can’t answer. When time was never priced in units, you never internalise its value. So a person earning the equivalent of $100/hour will still burn an hour to save $10 on a flight. And then the kicker:

Many Indian languages do not have a word for efficiency. And that’s true for many Asian languages. The word for efficiency does not exist. So then how do you value it if it’s not in a vocabulary?

Crocodiles, curiosity, and the predator who burns the fewest calories

Shah is a compulsive conjecture-machine: he dreams up a theory, then hunts for proof (lately via ChatGPT, which he calls his favourite toy). He asked it which animals survived 100+ million years nearly unchanged — horseshoe crabs, crocodiles — and what they share. Three traits, which he maps onto companies:

They can crank their metabolism down at will — survive a famine (or a COVID) by burning almost nothing, instead of burning cash and vanishing.
A very high conversion rate per hunt — they don’t chase everything; they wait, then deliver the deadliest bite. High judgment, not high activity.
They’ve weathered wild environmental change — they adapt, and adaptation comes from curiosity.

You don’t see a crocodile being busy. You see a crocodile just waiting patiently at the watering hole for that best meal. … The beautiful definition of a predator: the one who burns the least amount of calories to earn the most calories.

Curiosity, for Shah, isn’t cuteness — it’s a survival trait, and it requires security. Insecure people perform expertise; secure people happily ask “what does this word mean?” in a group chat of 60. He reframes wealth itself as information asymmetry: “all the best companies in the world have unfair information asymmetry,” built by endlessly collecting dots and connecting dots. He also runs a monthly ritual with his leadership team — what’s the hardest problem you solved last month? Most people, staying busy, have nothing to say. The extraordinary ones always do. A senior person’s real job, he says, is Chief Problem Solver.

Second-order thinking, and how you might build it in a child

Second-order thinking — judging the butterfly effects of an event, “if this happens, then B, then C” — is, Shah claims, the single best predictor of success. The brain hates it because it’s taxing, unless you trained it young to find the exercise rewarding. His interview question: “if everyone who took a COVID vaccine dies tonight, what happens to money, law, militaries, markets in 12 months?” Fewer than 10% of smart people answer well.

How do you grow it in kids? He has no children (“so I can run experiments on everybody else’s”), but he’s full of ideas. Strategy games build second-order thinking; physical games build discipline; both together is the jackpot. His “Wi-Fi school”: ask a child one why every meal — why do humans wear jewelry, why is Super Bowl advertising so expensive — and let them dig into history and origin stories. Why is it called a micro-phone? Going deep into origins forces the second-order muscle to grow on its own.

Contrarian corner: wealth is stored energy

Shah’s most heretical belief is that our understanding of wealth is simply wrong. Wealth, he argues, is stored energy — and because energy isn’t zero-sum, neither is wealth. Humans are the only species that learned to convert every form of energy (kinetic, fuel, solar, sound) to advantage, which is why wealth has shot upward since the Industrial Revolution and why AI and fusion will push it further. The corollary he knows is unpopular: wealth will always concentrate — “that’s the physics of wealth” — it merely changes mediums and countries. Chasing perfect equality, he says, fights physics; let people create wealth and there’s plenty to lift everyone. (He likens it to Elon Musk’s framing of wealth as just rows in a database you can expand infinitely.)

The gift of struggle

Shah’s life began in failure — a family financial crisis pushed him into work at 15, and he says he’s still, in some sense, running from it. He’s reluctant to dwell on it because he doesn’t want the “poverty card” sympathy. But it yields his sharpest line on parenting: the one gift successful parents can never give their children is the gift of struggle. Chips on shoulders, he notes, put chips in pockets. Entrepreneurs, meanwhile, have a “weird ability to forget about failures” while keeping the lesson — he claims he genuinely can’t remember COVID; it’s “just a blank in my head.”

Key Takeaways

Delta 4 test: rate the old solution and your new one out of 10. Unless the gap is ≥4, the product is reversible, un-brag-worthy, and won’t spread. A delta ≥4 product has near-zero CAC because users evangelize it.
A “tech product” can score worse than its analog alternative (suit-buying online). Being technological ≠ being more efficient.
ARPU is roughly a function of a country’s per-capita income. India = cheap DAUs, brutal ARPU (~$2,500/yr average income; Meta maybe earns $3–4/user/yr there).
Indian founders who copy the US “hundreds of millions of users” model must go abroad to find the ARPU to balance the equation, or die.
“Focus” is a curse in low-ARPU, low-trust markets. You must do many things, and weak institutions concentrate trust into a few mega-brands (Tata sells salt to cars). Super-apps are a symptom of low trust.
No Indian is paid by the hour, so the value of time is never internalised. Many Indian/Asian languages have no word for “efficiency.”
Rama/Krishna 2x2 (values × obedience): the best hired CEOs flip between rule-following Rama and rule-breaking Krishna while preserving the founder’s dharma rather than imposing their own ego. Nadella played Krishna during the OpenAI crisis, then reverted to Rama.
Brahma (create) / Vishnu (sustain) / Shiva (destroy) is a cycle every company runs — founders periodically must “play Shiva” and destroy to grow again (Zuckerberg’s reinvention).
Three traits of 100-million-year-survivor species (crocodile, horseshoe crab), applied to companies: dial metabolism down at will (survive lean times), high conversion-per-hunt (judgment over busyness), and adaptability via curiosity.
“The predator burns the least calories to earn the most.” Activity ≠ progress; ask monthly “what’s the hardest problem I solved?”
Wealth = information asymmetry; the best companies have unfair asymmetry built by collecting and connecting dots.
Second-order thinking is the top predictor of success; the brain hates it unless trained young via strategy games and “one why per meal.”
ChatGPT structurally rewards people who ask great questions and punishes those seeking basic answers — “the world is going to be unfair to people who can ask great questions.”
Status comes from demonstrating the ability to waste resources. Meta making its headset utilitarian and cheap handed Apple the premium-status crown.
Lab-grown diamonds will (like cultured pearls before them) become parasitic on diamonds’ status and destroy the profit pool — a worked example of second-order thinking.
A country’s profit pools reveal its values; copying another country’s profit pool is a mistake. India has few profitable retailers (low fashion spend, <1% divorce, low female labor participation); patriarchal societies skew market cap toward financial services over consumption.
The only gift a successful parent cannot give a child is the gift of struggle.

Claude’s Take

This is a genuinely good listen, and the rare founder interview where the frameworks survive contact with reflection. Shah’s strength is taking something everyone has felt vaguely — “this app isn’t catching on,” “India is weird to sell into” — and handing you a crisp tool to name it. Delta 4 and the DAU/ARPU asymmetry alone are worth the 90 minutes; both are portable far beyond India.

Now the BS filter. Shah is a professional aphorist, and aphorisms reward pattern-fit over truth. The crocodile traits, the Rama/Krishna 2x2, “wealth is stored energy” — these are post-hoc analogies, and he admits as much (“these are all conjectures which I can’t prove”). They feel profound partly because they’re unfalsifiable and beautifully delivered. The “no word for efficiency in Indian languages” claim is the kind of striking factoid that’s repeated precisely because nobody checks it (it’s shaky — Hindi has kshamta/dakshata). And there’s a built-in irony: a man who runs a credit-card app for India’s richest 25 million families is also the man assuring us that concentrated wealth is simply “physics” — a tidy philosophy for someone on the right side of it. Take the wealth-as-energy section as provocation, not economics.

But that’s the correct way to consume Kunal Shah: as a generator of sharp questions and reusable lenses, not received truths. He’d agree — his whole method is conjecture-then-verify, and he’s explicitly anti-guru (“the biggest profit-making scheme is telling people to love themselves”). High signal-to-noise for the genre, intellectually honest about its own speculation, and the India-specific material is the best short articulation of that market I’ve heard. An 8 — it loses points only where the elegant analogy outruns the evidence, which is often, but it tells you when it’s doing it.