Inside the Mind of Sanjeev Bikhchandani (Info Edge / Naukri)

ELI5 / TLDR

Sanjeev Bikhchandani built Naukri, took Info Edge public, and then used the IPO cash to become one of India’s most successful startup investors — backing Zomato and Policybazaar a decade before anyone thought they’d work. The twist: he insists he’s a bad public-market investor (he blew most of his savings on a stock tip in 1990 and stayed out for 26 years), and that his startup wins came from patience and not having a fund deadline, not genius. The interview is a tour of how he actually thinks about valuing companies nobody can value yet, what makes him walk away from a founder, and why holding Zomato for 15 years was only possible because his shareholders treated those bets as “cash on the balance sheet” and left him alone.

The Full Story

The investor who isn’t one

Bikhchandani opens by dismantling his own reputation. Asked how he became such an outlier returns generator, he says: “I never regarded myself as a great investor.” Then he tells the founding wound. In 1990, quitting his marketing job with 68,000 rupees in savings — one year’s post-tax salary — and “petrified,” a friend pointed him at a stock with a rumoured bonus issue. A sub-broker upgraded the gamble: buy on forward with margin, make 10x. He put in 60,000.

Number one, if you’re buying on a rumor, the rumor better be right. Number two, you better buy early in the rumor.

He was wrong on both. A week later he was down to 2,000 rupees. The lesson he drew was a circle-of-competence one before he’d ever heard the phrase: the stock market is not my area, stay away. He didn’t touch it again until 2016.

When he finally did return, it was less for money than to kill a chip on his shoulder — a listed-company founder who couldn’t talk shop at investor conferences. His method is the part worth keeping. He bought 30-40 investing books off Amazon, didn’t have time to read them, so read the tables of contents and blurbs. He noticed everything traced back to Warren Buffett and, before him, Benjamin Graham, and distilled four rules: invest in what you understand, buy quality, enter at the right price, hold for a long time.

Then he turned those into mechanical filters, because he didn’t trust himself to do research. Quality? Anything 30,000 crore market cap or above must be quality — “to get there is taking two three decades.” Knock out anything with a whisper on governance. Keep companies that 3x’d in five years, plus a few added back on reputation alone (Asian Paints, a couple of Tatas). That left ~30 names. Right price? Nobody knows, so don’t guess — run a weekly SIP across all 30 over three years. The result: 47% return against a benchmark of ~34%.

This is not pure tuka [fluke] by the way, this is pure first principles thing.

Then he read about a contracting balance sheet, got scared by the P/E ratios he was holding, and sold everything. Point proven, chip gone. He now hands public-market money to PMS managers and is happy with 17-20%. “I had to prove a point to myself that I’m not stupid.”

How the startup investing actually started: a compliance problem

This is the best story in the interview because it punctures the founder myth. Info Edge went public in 2006. Bikhchandani wrote the “objects of the issue” section of the prospectus himself — listing reasons like liquidity, ESOP value, giving early investors an exit. The banker told him none of that would pass SEBI; you have to say how you’ll use the money. So they wrote the standard boilerplate: organic growth, inorganic growth, general corporate purposes.

The problem: Naukri threw off cash and had negative working capital. They didn’t need the IPO money. Quarter after quarter the “use of IPO proceeds” line read nil.

By third quarter board got tense — guys, SEBI’s watching, you’re not using the money for what you said. It’d be embarrassing.

Nothing was worth buying in 2006. So someone reframed startup investing as a form of inorganic growth (you could acquire to majority later), the board approved a 160-crore budget over a few years — “don’t go overboard, don’t go crazy” — and one person was pulled off the business to do it as a side activity.

So out of a need for compliance we decided to start investing. It was not this great vision strategy.

Policybazaar came in 2008, Zomato in 2010. By 2012 they’d put ~250 crore into 8-10 startups and nothing looked like it was working.

The lemons ripen early

His sharpest line about early-stage investing:

We are in a business where the lemons ripen early. The blowouts come early, successes come after 7-10 years.

So the early years look bleak by design — and meanwhile he was a public-company CEO with institutional investors telling his CFO, behind his back, that “Sanjeev has gone mad,” that entrepreneurs are megalomaniacs who over-diversify, that he got lucky once and should stop wasting money. Under pressure, the board had him pause new investments from roughly 2012-2015 — exactly the window when Zomato and Policybazaar emerged as winners, and exactly when they missed several other good companies. Then the same investors complained he’d lost his mojo and stopped taking risk. He was unbothered, because — and this is the recurring theme — Info Edge is not a fund. No clock.

Why the no-deadline structure is the whole game

He frames the Indian exit problem cleanly. Inception to IPO in India typically takes 12-14 years or more (Info Edge was fast at ~9). Strategic sales at valuations a VC would enjoy basically don’t happen here outside a Flipkart. So a 20x — the kind you need one or two of to return a fund 4-5x — almost has to come from an IPO that arrives after most fund lives have ended.

All these VC funds which say they’re early stage — the truth is more than half their money has not gone in early stage. It goes in the later rounds, growth rounds. Because they’ve got to return LPs within the deadline.

Info Edge dodges this. Its funds are 12+2 years, not 10. It has two or three LPs (Temasek came in for half of Fund I, closed during COVID over a two-week phone call), so instead of being forced to sell winners it can offer LPs shares-in-kind. And the balance-sheet investments in Zomato and Policybazaar sit there for 15-18 years because analysts treat them as cash on the balance sheet and don’t pressure him — until an IPO appears on the horizon and the holdings suddenly become a sum-of-the-parts story worth scrutinising. He quotes Rakesh Jhunjhunwala on Titan: if you find a truly great company, hold forever.

How he reads a founder

A run of concrete tells. He does reference checks and is “very careful” if they’re bad. Running two businesses, or “I’ll go full-time the moment I get a cheque” — both are no-nos; not committed. His favourite trick question: what salary do you want to pay yourself? He doesn’t care about the number, only the reasoning. “40% more than my last job because I’m taking a risk” — wrong, you’re paid in equity for the risk. “Parity with my business-school batch” — wrong. “Enough to live on plus a bit, and I don’t want to raise too much because I value my stock and don’t want to dilute” — much better answer.

The deeper signal is whether the idea came from a customer insight. The Zomato story: Deepinder Goyal, then 28, running Foodiebay, was the only restaurant-listing site with all the menu cards — because he and his co-founders had ridden motorcycles to Delhi restaurants on weekends to scan them. “That’s foot slogging.” The origin was that as a Bain consultant, Goyal had scanned 80 restaurants’ menus onto his office page and within three days 95% of internal traffic was hitting it. The market had given him the signal.

His essential product feature was based on a customer insight — not “key Khurana is doing it, let me copy it.” His customer told him.

For founders who didn’t get the idea from a customer, he looks for “natural attraction” — downloads climbing week over week without ad spend. Some hook that gives it legs.

Valuing the unvaluable

How do you price a company before it has a PowerPoint? (He invested 20 crore in Policybazaar on a conversation, before a deck.) His framework is unusual:

We don’t look at the market size, we don’t look at the comp. We look at — what percentage of their lives do we deserve tomorrow? We are betting our money. The founders are betting their lives. We have a portfolio of investments. The founders don’t have a portfolio of lives. They have only one life.

Then the practical second question: after five to seven years of dilution, will the founders still own enough to stay motivated? If not, everyone’s money is at risk. So don’t get greedy on the first cheque.

He’s also blunt that projections are useless when a market is being created from scratch — and the surprises run upside. A VC once declined Naukri’s seed by arguing that converting the print recruitment-ad market to online would shrink it 100x, capping revenue at 4.5 crore. Info Edge today does ~2,500 crore in sales. When they raised from ICICI Ventures in the early 2000s doing 3 lakh/month, they projected 1 crore/month in four years and thought that aggressive — they hit 84 crore, wrong by 7x. “When new categories are being created, the market is surprisingly upside.”

The Zomato–Blinkit bet

Asked about Zomato buying Blinkit when both quick commerce and food delivery were loss-making and the stock was below its IPO price, Bikhchandani — on the Zomato investment committee — says they were nervous and he pushed back on Goyal directly:

Your stock price is down, investors unhappy, your main business is making a loss, Blinkit is making a loss. The only thing you have going is money in the bank.

Goyal’s argument was that the window was now — wait two quarters for food delivery to turn profitable and the opportunity is gone forever — and “trust me, I know how to turn around Blinkit.” They took a leap of faith. “Thus far at least they’ve got it right.” On verticalising quick commerce, he’s sceptical: a 10-minute dark store works because Blinkit sells 2,000+ items, so demand density supports the riders. A single-category store (“shoes in 10 minutes”) won’t have the demand density. He thinks B2B quick commerce and rethought same-day/next-day logistics are more promising than narrow 10-minute verticals.

Deep tech, AI, and buying what isn’t fashionable yet

His investing hypothesis in one line:

Go today into an area that’s not sexy or fashionable and invest, and it becomes sexy or fashionable after two years.

That’s how they entered deep tech around 2019 — handing a 35-50 crore balance-sheet corpus to Vibhor Sharma, a 2001-vintage Info Edge techie and former CTO who’d left for Amazon and come back. For two or three years the herd didn’t follow and they worried; now markups are “unimaginable.” The portfolio is genuinely eclectic: a vitiligo cure, drug discovery, waste-to-methane, electric planes, battery tech, Matter electric motorcycles, robotics. Deep tech, he notes, isn’t one technology — it’s 500 different ones across sectors that people bucket together.

On AI, he’s clear-eyed about India’s position: not LLMs (no compute, no GPUs, no capital, and the domestic market may not justify the spend) but the application/vertical layer, and above all data — 1.4 billion people. The play is to safeguard that data and ensure Indian companies own a share of any IP built on it. Inside Info Edge, 150 AI engineers work on non-trivial problems like surfacing the right CV out of 90 lakh résumés on page one instead of page fourteen, and voice bots that call the bottom 40-50% of clients who were never worth a human call.

Governance and institutions

Governance, he insists, has to be built in early because it becomes habit, and it “starts and ends in the founder’s head” — you have to want to be honest, pay full taxes, protect minority shareholders, and subordinate yourself to a board that can overrule you. The early conversations with ICICI Ventures in 2000 stuck with him: they forced a Big Five auditor, set his salary (zero at the time), and barred him from selling shares while they held. The PwC partner’s framing:

Your auditors and your independent directors are there to save you from yourself. We are on your side, we’re not policemen.

On the tough periods: post-dotcom, with 16 months of cash and a salary bill 4x revenue, he personally did 64 sales calls in one month — “founder mode, as they call it now.” During COVID, with revenue down 44%, stress-testing showed three years of runway even at zero revenue, so they sacked nobody and instead turned the call centre into a COVID-relief operation for employees and families. He refuses the hero narrative: “If Naukri wasn’t there, if it wasn’t 93% gross margin, negative working capital, cash in the bank, I don’t pass this test.”

Info Edge is deliberately not a family business — “you can’t be in tech and be a family business.” He doesn’t think about legacy, only about building institutions that outlive him; nothing is named after him despite being a large donor (Ashoka University). His closing north-star metric: “Tomorrow if I vanish, what will happen? Best [if it] works without me.”

Key Takeaways

The lemons ripen early. In early-stage investing, failures surface in 1-2 years; successes take 7-10. The early portfolio always looks bleak — that’s the shape of the asset class, not a verdict on it.
No deadline is the edge. Info Edge held Zomato 15 years, Policybazaar 18, because it isn’t a fund with a 10-year clock. India’s inception-to-IPO is 12-14+ years and strategic exits barely exist, so deadline-driven VCs are structurally forced into later, safer growth rounds.
Price the founder’s life, not the market. Ask what percentage of a founder’s one life your money deserves — and whether, after years of dilution, they’ll still own enough to stay motivated. Greed on the first cheque destroys alignment.
The salary question is a character X-ray. “What will you pay yourself?” — the reasoning (“enough to live on, don’t want to dilute”) matters far more than the number.
Customer insight beats copying. The best ideas come from a founder’s own lived pain (Goyal’s menu cards). Absent that, look for “natural attraction” — organic download growth with zero ad spend.
Buy what’s unfashionable; let the herd arrive later. Enter an unsexy area now, let it become sexy in two years — you bought cheap and the markup follows.
New categories surprise on the upside. When a market is created from scratch, projections are near-useless and usually too low (Naukri beat its four-year projection by 7x; a VC capped it at 4.5 crore — it’s now ~2,500 crore).
Quick commerce verticalisation needs demand density. 10-minute delivery works because a dark store carries 2,000+ items; single-category “shoes in 10 minutes” can’t support the riders.
Governance is a habit set in the founder’s head. Build it early or you acquire bad habits that are hard to shake. Auditors and independent directors “save you from yourself.”
Buffett, arrived at by accident. Invest in what you understand, buy quality, enter at the right price, hold long — Info Edge backed into all four before reading the theory, then mechanised them into filters for public markets.

Claude’s Take

This is a better interview than most founder profiles because Bikhchandani actively refuses the flattering version of his own story. The “we started investing out of a SEBI compliance problem” anecdote is the tell — a lesser interviewee would have retrofitted a visionary thesis. He keeps doing this: attributing wins to structure (no fund deadline, 93% gross margins, negative working capital) rather than insight, and crediting his team (Kitty Agarwal reversing his Policybazaar sell, Vibhor Sharma on deep tech) at every turn. That’s either genuine humility or a very sophisticated form of it; either way the structural points survive the skepticism. The no-deadline argument in particular is the most useful thing here — it’s a real, non-obvious explanation for why Info Edge’s track record is hard for conventional VC to replicate, and it generalises.

Where to keep one eye open: survivorship and selection. We hear the wins (Zomato, Policybazaar, Naukri) narrated with their happy endings known; the 8-10 startups from 2007-2012 that didn’t make it get one line (“the lemons ripen early”) and no autopsy. The “47% beat the benchmark” public-market story covers a roaring 2018-2021 bull window and a clean exit before the drawdown — proving the process beat the index over one favourable cycle isn’t quite proving the process. And the Blinkit section reads as vindicated mostly because it worked; “trust me” from a founder is exactly the input his own framework warns against, and he knows it.

Docking it slightly off a 9 only because the interviewer mostly lobs and rarely presses on the failures or the selection bias, so we get the curated version. But the frameworks are genuinely portable and stated plainly, the stories are concrete and load-bearing, and the self-awareness is real. 8/10.