There is a particular kind of startup failure that nobody in the ecosystem quite knows how to talk about. Not the early-stage misfire — the two-person team with a slide deck and an idea that never found a market. Not the pandemic-era casualty, swept away by forces that no founder could have predicted. The kind of failure that happened in India across 2025 is more instructive, and more unsettling, than either of those.
These were companies that had raised hundreds of crores of investor money. They had real users, real revenue, and real operations spread across multiple cities. Several were valued at over a thousand crore rupees. A few were unicorns. And they collapsed anyway — sometimes spectacularly, sometimes in slow motion — in ways that reveal deep structural problems the Indian startup ecosystem has been too polite to acknowledge directly.
According to data from startup intelligence platform Tracxn, 729 startups shut down in 2025. That number is significantly lower than the previous year’s tally — a fact the ecosystem has used to project cautious optimism. But the raw count obscures something important: the 2025 closures were disproportionately concentrated among the most prominent names. The companies that failed this past year were not the ones that should have failed. They were the ones the industry had held up as examples of what Indian entrepreneurship could build. Understanding why they fell — and understanding it precisely, rather than charitably — is the most useful thing anyone building or investing in a startup in India can do right now.
The Startup Names We Lost
BluSmart will likely be remembered as the defining startup failure of this era — not because it was the largest, but because the circumstances of its collapse exposed so many things at once. Founded in January 2019 by brothers Anmol and Puneet Singh Jaggi and their co-founder Punit K. Goyal, BluSmart entered India’s ride-hailing market with a proposition that felt almost too good to be true: a 100 percent electric fleet, salaried drivers (rather than commission-based contractors), zero ride cancellations, and fixed pricing with no surge.
It was, by all metrics, the kind of company that sustainable mobility advocates had been asking for. At its peak, it operated over 8,700 EVs across Delhi-NCR, Bengaluru, and Mumbai, commanded roughly 9 percent market share in Delhi, and had raised approximately ₹1,394 crore — close to $168 million — from credible international investors including BP Ventures and the Swiss impact fund ResponsAbility.
What actually destroyed it had nothing to do with its product. SEBI’s investigation into Gensol Engineering — the solar EPC firm also promoted by the Jaggi brothers, which owned the majority of BluSmart’s EV fleet and leased it back to the startup — found that approximately ₹262 crore in loans sanctioned for EV procurement had been diverted for personal expenditures: luxury apartments in Delhi, foreign travel, golf equipment.
The regulator barred the founders from the securities market in April 2025. Within days, the company suspended all ride bookings, asked its 10,000 driver-partners to return their vehicles, and left around 500 employees without salaries. By July 2025, the National Company Law Tribunal in Ahmedabad admitted an insolvency petition. Users with money in BluSmart’s closed wallet were told, per the company’s own terms and conditions, that the funds were “strictly non-refundable.”
The deepest lesson of BluSmart is not about corporate fraud — that happens everywhere, and always will. It is about a structural vulnerability that was hiding in plain sight. BluSmart did not own its own fleet. It was operationally dependent on a separate listed entity run by the same founders. The moment that entity came under regulatory scrutiny, the entire business model collapsed instantly, with no fallback and no runway. BluSmart’s investors, including sophisticated international funds, apparently did not identify this linked-entity exposure as a critical risk. That oversight is as instructive as the fraud itself.
Dunzo is a different and in some ways sadder story — not one of fraud, but of a company that genuinely tried everything and still couldn’t make the numbers work. Founded in 2014 by Kabeer Biswas and three co-founders as a WhatsApp-based errand service in Bengaluru, Dunzo spent years becoming something genuinely beloved in urban India. The phrase “just Dunzo it” entered casual conversation the way “Google it” had a decade earlier. The company attracted over $449 million in funding over its lifetime, including Google’s first direct investment in an Indian startup (₹76 crore in 2017) and a landmark ₹1,645 crore injection from Reliance Retail in January 2022, which gave the conglomerate a 25.8 percent stake.
The pivot that killed Dunzo was the decision to enter quick commerce in August 2021, launching Dunzo Daily — a 19-minute grocery delivery service built on a dark store network of over 130 warehouses across Indian cities. The quick commerce bet required massive, sustained capital expenditure: dark stores are expensive to set up, expensive to staff, and require extremely high order density to become profitable at the unit level. Dunzo’s losses surged to a staggering ₹1,801 crore in FY23, while its operating revenue for that year reached only ₹226 crore.
The ratio of losses to revenue was not a growth-stage investment story — it was a structural indictment of the economics of the model. Blinkit, Zepto, and Swiggy Instamart all had deeper pockets and clearer paths to density in their core markets. Dunzo was competing against better-funded rivals for a market that couldn’t sustain all of them. By early 2025, the app and website had gone dark. Reliance formally wrote off its ₹1,645 crore investment. Hundreds of delivery workers and employees were left without wages.
The Good Glamm Group represents a third kind of failure entirely: the collapse of a thesis. Founded originally as MyGlamm in 2015 by Darpan Sanghvi, the company raised $342 million over 22 rounds from investors as distinguished as Warburg Pincus, Naspers, Amazon, Accel, Bessemer, and Prosus Ventures, achieving unicorn status with a valuation of $1.25 billion. Its strategy was India’s most ambitious attempt at the “Thrasio model” — the idea of building a house of brands by acquiring direct-to-consumer companies and converting their content audiences into commerce at scale.
Between 2020 and 2022, Good Glamm acquired over eleven companies, including The Moms Co., Sirona Hygiene, St. Botanica, Organic Harvest, and the digital media properties ScoopWhoop, MissMalini, and POPxo. The logic was that owning the content would reduce customer acquisition costs and create a distribution flywheel.
The logic never worked in practice. Many of the acquired brands were either loss-making or too small to justify the prices paid. The promised synergies — centralised marketing, shared supply chains, content-to-commerce conversion — failed to materialise at anything close to the projected scale. Revenue reached ₹603 crore in FY23 but losses that year were ₹917 crore. By early 2025, board members from Accel, Prosus Ventures, and Bessemer had all resigned.
The company began unwinding its acquisitions at brutal discounts: Sirona, acquired for ₹450 crore, was sold back to its original founders for ₹150 crore; ScoopWhoop, bought for ₹100 crore, was sold for ₹20 crore; MissMalini went for approximately ₹4 crore. By July 2025, CEO Darpan Sanghvi publicly acknowledged that the strategy had failed, writing in a LinkedIn post that Good Glamm had tried “too much, too fast, too big” and described momentum as “intoxicating — until you drown in it.”
Hike is perhaps the most poignant case on this list because its final collapse was largely caused by something beyond founder control. Founded in 2012 by Kavin Bharti Mittal — son of Bharti Airtel’s Sunil Mittal — Hike was once a genuinely exciting challenger to WhatsApp and Facebook Messenger, reaching over 100 million users at its peak, achieving unicorn status, and raising more than $250 million from global investors including Tiger Global, SoftBank, and Tencent.
By 2021, the messaging era had passed and Hike pivoted, shutting down its messaging app and launching Rush, a real-money gaming platform offering casual games like Ludo and Carrom. For a few years, it worked reasonably well. Then in 2025, the Indian government passed the Promotion and Regulation of Online Gaming Act, which imposed a blanket ban on real-money gaming platforms. Rush’s entire revenue base disappeared overnight. Hike explored pivoting to international markets — the US, the UK, Canada, Australia — but the founders eventually concluded that a full global reset would require more capital and time than was justified, and wound down operations. The company returned its remaining cash to investors. It was, of all the failures on this list, the most honourable exit.
The Pattern Nobody Wants to Admit
Taken together, these four stories are not isolated misfortunes. They are four different expressions of the same underlying problem, which is this: a significant portion of India’s celebrated startup ecosystem was built on a foundation of subsidised growth — capital standing in for genuine consumer willingness to pay.
The 2021 funding boom flooded the ecosystem with capital at a pace the underlying businesses could not absorb responsibly. Dunzo raised $200 million from Reliance at a $775 million valuation in January 2022, precisely as the funding winter was about to begin. Good Glamm raised its largest round in December 2021 at the absolute peak of venture valuations. BluSmart was planning a ₹425 crore funding round that fell apart in early 2025 when investor confidence evaporated. All three had burned through capital at rates that were only sustainable if the next round arrived on time and at an upward valuation. When the funding cycle turned, the model was exposed.
The deeper structural issue is about sectors. Quick commerce, EV fleet mobility, and content-to-commerce roll-ups all require enormous capital outlays before a single rupee of genuine profit is possible. That is not inherently disqualifying — infrastructure businesses always do. But the consumer pricing in these models was systematically set below cost, subsidised by investor money, to drive adoption.
Dunzo burned approximately ₹230 per order on its quick commerce operations at its peak. BluSmart’s economics depended on a leasing arrangement with a related party rather than genuine asset ownership. Good Glamm’s content-to-commerce conversion rates never justified the acquisition premiums paid. In each case, there was a version of the spreadsheet that worked — provided you believed in a set of projections about future scale that the actual market never validated.
There is also a governance crisis embedded in these failures that the ecosystem has been reluctant to name plainly. BluSmart’s founders were allegedly diverting funds from a listed entity into personal luxury purchases while their startup was burning cash and their employees were going unpaid. Good Glamm’s audited financials had not been filed for two years before its collapse.
The pattern of delayed salaries, unpaid vendors, and investor board members departing in silence before a public acknowledgment of crisis was repeated across multiple companies. When the governance structures around a company are weak enough that fraud or opacity can survive for years, the investors and board members who were supposed to provide oversight carry real responsibility for what follows. The ecosystem’s tendency to treat every major founder as a visionary beyond ordinary scrutiny is itself a structural risk.
What the Survivors Are Doing Right
The contrast with the startups that are thriving in 2026 is instructive precisely because some of them are operating in the same sectors as those that failed.
Zepto is the most striking example. Founded in April 2021 by then-teenagers Aadit Palicha and Kaivalya Vohra, it operates in exactly the same quick commerce space that destroyed Dunzo. In October 2025, it raised $450 million in a Series F round, pushing its valuation to $7 billion. Its revenue surged to over ₹11,000 crore in FY25 — a 149 percent year-on-year increase — while its per-order unit economics improved consistently. The critical difference between Zepto and Dunzo is not the model; it is the execution discipline.
Zepto focused intensely on achieving order density in a concentrated set of high-demand urban neighbourhoods before expanding. It kept governance clean and its cap table uncomplicated. It resisted the temptation to run IPL sponsorships as a substitute for operational efficiency. When Zepto paused its IPO plans in mid-2025 for six months to focus on reducing cash burn and improving profitability before relisting, it demonstrated the kind of patience and accountability that Dunzo’s leadership lacked.
Meesho’s IPO in December 2025 — at a valuation of roughly ₹53,000 crore on revenue of ₹9,900 crore — tells a similar story. What Meesho got right was its market identification. While other e-commerce players fought intensely over metropolitan India’s digitally sophisticated consumers, Meesho spent years building trust and logistics in Tier 2 and Tier 3 India — a market that was underserved, less price-sensitive than assumed, and enormously large. The company’s business model, built on a social reselling model that used WhatsApp-enabled word-of-mouth rather than expensive paid advertising, kept customer acquisition costs structurally low. The result was a company that grew without burning capital at Dunzo-scale rates.
Razorpay offers perhaps the most sobering lesson of all. In FY25, it reported a net profit of approximately $48 million on revenue from processing over $180 billion in annual payments. This is a B2B infrastructure business — not glamorous, not a consumer brand, not IPL-sponsored — built patiently over 11 years by Harshil Mathur and Shashank Kumar. Its survival and success in 2025 while peer-to-consumer businesses burned and folded around it is a quiet but emphatic argument for the proposition that sustainable unit economics are not a constraint on ambition — they are the foundation of it.
The common thread across the survivors is not luck, sector choice, or even product quality. It is the refusal to treat the capital raise as the product. The companies that are still standing in April 2026 built businesses where consumer behaviour, not investor subsidy, drove revenue. They kept their governance clean and their dependencies transparent. They expanded only when the existing base had earned the right to expansion. These are not glamorous disciplines. They do not make for good launch press releases. But they are what real companies are built from.
India’s startup ecosystem in 2026 is genuinely one of the most exciting in the world. Twenty-eight tech startups went public in 2025 alone — nearly double the number from 2024. Capital is returning, selectivity is improving, and a generation of second-time founders who carry the lessons of these failures into their next ventures is already at work. The graveyard is real, but so is the ecosystem growing around it.
The question worth asking, as you read this, is a simple one: Would you trust an app again if it was from a founder whose previous startup failed? The answer, increasingly, should probably be yes — but only if they can tell you precisely what went wrong, and show you, concretely, what they did differently this time. That is what the ecosystem’s next chapter requires: not the erasure of failure, but the honest accounting of it.
