产品要比旧办法高出四分，人们才会停不下来地讲

Shah describes himself as the only humanities and philosophy major founder in Indian tech. He wondered why he succeeded with his first startup, Freecharge, which he exited in 2015 for nearly $450 million, when he would never have qualified for the elite IIT cohort. That question sent him on a philosophical search for what makes things succeed, and the result was the Delta 4 framework.

The idea is simple: ask people to score the efficiency of the old way and the new way out of ten. Lenny Rachitsky rates a traditional cab three and Uber nine. When the gap is four or more, Shah says, three things happen: the change is irreversible, users forgive occasional failure, and a "Unique Brag-worthy Proposition" appears, so people cannot stop telling others about it. That is why Delta 4 products tend to have low or zero customer acquisition cost, as with Google, which spread through demos rather than ads.

Making something tech-based does not make it more efficient. Shah uses buying a suit online versus offline as the counterexample: Rachitsky rates online five and offline seven, so the gap is below four, the behavior stays reversible and nobody brags. Shah says the framework applies to features, products and whole businesses, and that he traced it to ideas from entropy and evolutionary biology.

Rachitsky says he has shared it with at least 30 founders as a way to explain why a product is not taking off. A product that is merely better than Excel does not clear the bar, because busy people will not dig into something that is only slightly better. He adds that the rating has to be four points higher than the current solution.

Rachitsky notes that the CEOs of Microsoft, Alphabet, Adobe, IBM, Palo Alto Networks and Starbucks were all born in India and emigrated to the US, and asks what can be learned from this. Shah offers conjectures he stresses he cannot prove. Immigrants who leave humble backgrounds have a chip on their shoulder and a hunger born of struggle, and Indian society's early emphasis on math and engineering creates status around hard exams.

His more distinctive explanation comes from author Devdutt Pattanaik's use of Indian mythology in management. Brahma creates, Shiva destroys and Vishnu sustains, and sustaining is the harder job. On a grid of values versus obedience, Krishna is high on values and low on obedience, the entrepreneurial type, while Rama is high on both and follows principles. Shah calls Ravana and Duryodhana the types you do not want to be.

Many CEOs do well, he argues, because they keep the founders' dharma, saying the principles were given to them and they will make them bigger. He says Tim Cook is maintaining the dharma of Steve Jobs, and that Satya Nadella played Krishna during the OpenAI crisis before returning to Rama. Chasing a personal legacy by changing the logo and the name is where creative destruction begins. Rachitsky connects this to Jim Collins's level five leaders.

Shah doubts these executives think about mythology consciously, but believes they would agree because they grew up with long-term value systems. He links this to a long-term cultural orientation, citing arranged marriages and a divorce rate below one percent, and argues that bad behavior in humans comes from short-term thinking.

Rachitsky raises the claim that around 90 percent of startups in India fail, with higher personal stakes. Shah responds that long-term societies are naturally risk-averse, and that great CEOs do not automatically make great founders because they have become sustainers rather than creators. He notes that things are changing: founders are respected, unicorns are celebrated, the Prime Minister speaks about startups and there is a National Startup Day.

The core human behavior has not changed, though. Because arranged marriage is still common, a failed founder can find it harder to marry. Shah gives the example of a manager at a large consumer goods company who launches a zero-to-one product that fails while a peer on a stable brand rides natural tailwinds and is promoted. India, he says, does not yet celebrate risk-takers.

To show what a different attitude looks like, he recalls a church in Portugal where only royalty and explorers like Vasco da Gama were buried. When a country gives its highest status to risk-takers, he argues, people respond, since everyone is seeking markers of social validation.

Shah explains why daily users are easy to win in India and revenue per user is not. ARPU depends on per-capita income, and he cites about $2,500 a year, so a product cannot extract $100 per user. Global companies come for cheap data and high smartphone penetration, and he guesses that Meta earns only a few dollars per Indian user per year. India becomes a great monthly-active-user farm, and Indian founders who copy Western plans of hundreds of millions of users end up having to find revenue abroad.

He says expectations that Spotify, Netflix or Amazon Prime would see tens of millions of Indians pay as soon as they arrive have not panned out, because free data, free content and short video compete for attention. Rachitsky's takeaway for investors is that 100 million users in India are not the same as 100 million in the US. Shah adds that no Indian has ever been paid an hourly wage, so most people cannot say what an hour of their time is worth, and some Indian languages have no word for "efficiency".

Shah argues that focus can be a curse in these markets, partly because revenue per user is low so companies must do many things. He also says low-trust markets have weak institutions, so consumers are wary of new things and trust concentrates in a few brands and super apps. Tata sells everything from salt to cars, and in India trust still often comes from the name of a person or family behind a business.

Asked about India's biggest challenges, Shah says that challenges and opportunities are often the same. Low female labor participation could be reshaped by AI and remote work, low per-capita income could be helped as AI spreads expertise, and he jokes that the largest employer in the world is inefficiency. He suggests interviews should include tasks only someone good at AI can complete, since people who cannot leverage it will become a liability within five years.

He notes that India's young generation is the first to truly learn entrepreneurship and risk, and layoffs at startups become national news. CRED's insight, he says, was that income and the value of time are concentrated in about 25 million families who are more global than the rest of India, so he built only for them. Investors wanted the next China, but only the population was similar. His prior success helped him raise a $25 million Series A.

The lessons since: companies great at zero to one will not naturally be great at ten to a hundred, and the founder has to evolve. Entrepreneurs are "uncertainty absorbers" for employees, investors and customers, but a growth investor or sovereign fund expects far more stability than a seed investor. Shah says Mark Zuckerberg had to play Shiva over the past few years to make Meta big again, and that the Brahma, Vishnu and Shiva cycle is universal.

He adds that a country's profit pools reveal what it values, so copying another country's pools is unwise. India has very few retailers among its thousand most profitable companies, while low divorce rates, arranged marriages and low female labor participation mean women's fashion spending is lower than men's, unlike elsewhere. He observes that patriarchal societies often have large financial-services market caps rather than consumption ones.

Rachitsky says people on Twitter asked most often when CRED will become profitable. Shah explains that India's business tradition is trading: buy low, sell high. Spending on distribution first and monetizing later is unfamiliar, and even his own father did not respect him until he built a profitable company. Tech makes up around 27-30 percent of US market cap but probably two to three percent in India, so he calls his the first generation that must prove it can build very large profitable tech companies.

He notes that investors have asked him about profitability far less than Twitter users have, which creates an arbitrage for founders who can prove a model with fast revenue growth. He also describes endless trolling in Indian comment sections, which he explains with the idea that envy is hyperlocal: people do not envy Elon Musk but do envy a local founder who was like them a few years ago.

His advice to founders facing hate is to take criticism seriously only from people who have done better than you and understand the nuance, while still welcoming customer feedback. He jokes that elements with low valency are called noble gases because they are the hardest to get a reaction from. Despite all this, he calls India the most promising market to build in, thanks to digital public infrastructure, government support and an evolved ecosystem.

His one regret is never having trained with the best product minds. He wishes there were an "operation theater" for watching Brian Chesky make a product decision, Mark Zuckerberg kill a product, or Tim Cook reject ideas before a launch, since many Indian founders had to figure everything out alone.

Shah defines a curious person as someone who is not proud of their expertise and gets excited when facing a problem they cannot solve. Curiosity requires security, he says: a CEO should feel fine asking a dumb question in a company WhatsApp group, while insecure people try to display expertise or quietly search instead. That slows compounding growth and the ability to adapt.

He shares an exercise he did with GPT, asking which species have survived over 100 million years without changing much. Sharks, horseshoe crabs and crocodiles share three traits: they can slow their metabolism at will, they convert a very high share of their attempts to get food, and they adapt to environmental change. He maps these onto companies in COVID, which either slowed burn and survived or burned out, and onto people who either retreated or got curious.

For Shah, wealth is information asymmetry, and the best companies have an unfair share of it that comes from constantly collecting and connecting dots. Rachitsky links this to restaurateur Danny Meyer's "always be collecting dots". Shah says he avoids naming favorites because the word limits the mind, and he asks his direct reports every month what the hardest problems they solved were. A senior person's role, he says, is to be the chief problem solver, and like a crocodile, to burn few calories for the best meal.

His learning method is to form a conjecture, then use GPT and wide research to look for proof across fields like chemistry, physics and human behavior. He gives lab-grown diamonds as a recent obsession, comparing them with cultured pearls that lost their royal status once anyone could have them. He thinks GPT will favor those who ask great questions, and that every book read makes the brain poorer for the next.

To ask great questions, Shah suggests studying people who repeatedly make the right choice among many, such as hiring great people or choosing a great partner, and trying to extract how they decide. They often cannot articulate their mental models. The other question to ask is about second-order effects: what will happen to the world and why do they think so.

He considers second-order thinking one of the best predictors of success and wonders what in childhood builds it. His loose finding is that physical games build rigor while strategy games build second-order thinking. He recommends parents run a "Whyfi school", asking one why-question at each meal, such as why humans wear jewelry or why Super Bowl ads are so expensive, and letting children research it overnight.

Using AI to answer such questions is fine, in his view, as we are all becoming co-pilots of AI. The point is to derive the second-order insight and ask where else the pattern appears. In his interview question he asks candidates what the world looks like in 12 months if everyone who took a COVID vaccine died tonight, and says fewer than 10 percent of very smart people answer it well.

He defines second-order thinking as judging the butterfly effect of an event as accurately as possible, and says the brain hates it unless trained early to find it rewarding. In an answer he looks for range and the leaps a person can make from A to B to C.

In the contrarian corner, Shah says our understanding of wealth is flawed: wealth is stored energy, and since humans convert every form of energy to their advantage, wealth is not zero-sum. He argues equality of wealth will never exist and chasing it is a bad idea, because letting people pursue wealth creates enough for everyone else. Rachitsky shares Elon Musk's database metaphor, and Shah adds that the database can grow without limit.

On failure, Shah says he has had many: failed products, hires, broken trust and betrayals. Entrepreneurs, he says, forget the story of a failure but keep the lesson, and life is too short to make all the mistakes yourself. His own life began with a failure: his family went through a severe financial crisis and he started working at 15, and he says he is still escaping it. He only recently began talking about it, not wanting sympathy, and calls the gift of struggle the one thing successful parents cannot give their children.

In the lightning round he recommends books on human behavior and evolutionary biology, and says he does not re-read books because repetition becomes too religious. He loved Oppenheimer and is waiting for his Vision Pro, arguing that Zuckerberg's pitch of utility actually helps Apple, since status goes to products that signal waste rather than efficiency.

His closing motto is that we are transient and should collect as many stories as possible. His "Monday Motivation" posts are meant as anti-motivation, a reminder that success comes from pain and hard work with no shortcut. He says the best way to love yourself is to keep evolving.