By Anish Saraf , Investment Professional , Lighthouse Canton 

In 2021, India was minting a new unicorn roughly every eight days. Forty-four companies crossed the billion-dollar mark that year, more than the previous three years combined, and startup funding touched $42 billion. I watched that year from a business school classroom, where the case studies preached unit economics while the market outside ignored them. 

Then the tap closed. Two new unicorns in all of 2023. Funding down to roughly $10 billion. Even in 2025, with about $11 billion raised, only six companies turned unicorn. The number that once defined ambition has quietly become a lagging indicator.

It would be easy to read this as a mood swing: capital got expensive, everyone sobered up, and the cycle will eventually swing back to the frenzy. I don’t think that’s what happened. The exit market itself changed, and that makes the correction structural.

Look at the scoreboard the public market keeps. Not the famous cautionary tales, but the quieter pattern underneath. Groww came to market as a genuinely profitable fintech and roughly doubled inside eight months. ixigo went public already making money, listed 78% up on debut, and has held around twice its issue price since. Policybazaar spent its early listed years as a punching bag, until it strung profitable quarters together and the market re-rated it hard. Now the other side of the ledger. Ola Electric listed to enormous fanfare, briefly multiplied, then bled to roughly half its issue price as losses widened and volumes slid. Mobikwik popped nearly 60% on debut and slipped below issue within months once the losses returned. Track the whole listed new-age cohort and the split is stark: by late 2025, the profitable companies had risen a median of 31% while the loss-makers had fallen 42%. That is not sentiment. That is repricing. And once the IPO window demands positive contribution and a credible path to profit, the discipline cascades backwards through the private markets, into Series C, then B, then A.

Here is the part I find most telling, though, and it comes from my own seat. I entered the industry once the correction was already underway, on the equity side of venture capital; I now work in venture debt. For lenders, unit economics were never optional. A loan carries no upside optionality. There is no 100x winner to paper over the misses. Credit gets extended only when a business is predictable enough to service it. Which is why the growth of venture debt in India is itself evidence of maturation: from a niche product a few years ago to $1.23 billion across a record 238 deals in 2024, and growing again in 2025. Credit scales in an ecosystem only when the underlying businesses become legible enough to underwrite.

The lender’s checklist is short, and equity investors increasingly ask the same questions:

• Contribution margin after all variable costs ; not CM1 theatre. If a unit loses money once delivery, returns and support are loaded in, scale just multiplies the loss.
• CAC payback :  How many months before a customer repays what it cost to acquire them? Under twelve is comfortable; past eighteen is a runway problem.
• Burn multiple: How much cash do you burn for every rupee of net new revenue? It is the bluntest measure of what growth actually costs.
• Cohort retention. Do older customers stay and spend more, or is growth just fresh acquisition masking churn?

None of this is an argument against growth. Some businesses genuinely need scale before the economics resolve; network-effect platforms and capital-heavy infrastructure earn their margins only at density. The discipline that matters is sequencing: prove the unit works, then pour capital on it. The 2021 playbook got the order backwards, funding scale in the hope that economics would follow.

They rarely did. PharmEasy went from a $5.6 billion valuation to raising capital at a 90%+ discount, after breaching a loan covenant on its own debt. That is what growth without a working unit eventually compounds into.

The funding winter didn’t invent these metrics. It just made them unavoidable. Unicorn status, it turns out, is an output of unit economics, not a substitute for them.

                                                                                                                                                                      Coordinated by: Kashish Srivastava