Zepto Case Study: How Two Teenagers Rewired India's Grocery Market

Two 19-year-olds with no retail experience, no warehouse, and no delivery fleet built a business that posted ₹9,669 crore in revenue in its fourth full financial year. That is not a typo. Zepto — co-founded by Aadit Palicha and Kaivalya Vohra in July 2021, shortly after both dropped out of Stanford University — grew its top line 129% year-on-year in FY25, handling roughly 1.5 million grocery orders per day across India's largest cities by March 2025. The losses are real too (₹3,367 crore in FY25, up 177% from the prior year), and we will get to those. But the growth story, and the operational logic behind it, is one of the most instructive case studies in Indian startup history.

The common narrative around Zepto is that it sold a promise — "10 minutes or less" — and that the promise was the product. That reading misses the real breakthrough. Zepto's durable advantage is not the clock; it is the real estate. By building a network of 1,000+ micro-warehouses — dark stores — in the densest residential neighbourhoods of Mumbai, Bengaluru, Delhi, Hyderabad, and Chennai before any competitor understood what they were doing, the company locked up the physical infrastructure that makes 10-minute delivery possible. Every competitor who now wants to match that speed must replicate not code, but warehouse leases, cold chain logistics, and neighbourhood-by-neighbourhood SKU selection — a process that takes 12 to 18 months per city. That is the actual moat.

From KiranaKart to Zepto: The Origin in Three Pivots

The story begins not in 2021 but in 2020, during India's first COVID-19 lockdown. Palicha and Vohra — childhood friends from Mumbai who had both been admitted to Stanford's Computer Science programme — found themselves stuck at home, watching elderly neighbours struggle to get groceries delivered in under two to three days. They started delivering by bicycle themselves, sourcing from the kirana down the street and charging nothing. They called it KiranaKart.

The first insight was logistical, not technological: the neighbourhood kirana had the right products but the wrong fulfilment model. It was built for walk-in customers, not timed delivery. Orders came in on WhatsApp, stock wasn't digitised, and a single delivery person couldn't handle more than 8–10 drops per hour. KiranaKart's model of outsourcing picking to existing kiranas hit a ceiling quickly — you inherit the kirana's inefficiencies along with its stock.

Pivot one came when they convinced Contrary Capital to fund them if they dropped out of Stanford. With fresh capital, they rebuilt the model around company-controlled inventory — small warehouses that only Zepto staff could access. No customers walking in. No distractions from counter sales. Every square foot of the store optimised for order picking. They renamed the company Zepto and applied to Y Combinator, which backed them in 2021.

Pivot two was geographic focus. Rather than spreading thin across India, Zepto opened 10 dark stores in Mumbai and went deep — learning SKU velocity, optimising routes, training pickers to achieve a 3-minute pick time. The operational learning from those 10 stores became the playbook for every subsequent city. By Q4 2021, Zepto had raised $60 million at a $225 million valuation, and the dark store count was already at 40 across two cities.

Pivot three — and the one that separated Zepto from every earlier grocery delivery attempt — was the decision to stock just 8,000 SKUs per dark store rather than the 30,000–50,000 SKUs of a BigBasket or an Amazon. This is counterintuitive: fewer SKUs sounds like a weaker offering. But every additional SKU adds warehouse space, picking complexity, and demand forecasting difficulty. Zepto's 8,000 SKU selection covers 85–90% of a typical urban Indian household's monthly grocery spend. The remaining 10–15% — niche imports, specialty items, bulk buys — is where the company deliberately chose not to compete. It is a deliberate constraint that makes everything else possible.

The same principle of deliberate constraint applies to D2C brand building in India. Our earlier analysis of how Indian D2C brands get their first 1,000 customers found that the founders who grew fastest were the ones who picked one channel, one customer type, and one geography — and went deep before going wide.

The Dark Store Model: Inside the Machine

A Zepto dark store is a 2,000–3,000 square foot unit — roughly the size of a large urban flat — located in a high-density residential neighbourhood, within 1.5 to 2 kilometres of the customers it serves. The location is everything: Zepto's average delivery time of 8 minutes and 47 seconds is a function of proximity, not speed. Its riders are not especially fast; the stores are just close.

Each store stocks approximately 8,000 SKUs — fruits, vegetables, dairy, snacks, packaged staples, personal care, and basic household goods. Products are arranged not by category but by velocity: the 500 fastest-moving items are placed closest to the packing station, reachable without walking more than three metres. A trained Zepto picker can locate, pick, and pack an order in under three minutes. The remaining 8 minutes of the delivery window are spent on the last-mile ride.

The staff model is lean: 10 to 15 associates per store across three shifts, plus a store manager. Each store runs roughly 900 to 1,200 orders per day when fully ramped. At an average order value (AOV) of ₹625 to ₹650, that is ₹56 to ₹78 lakh in gross merchandise value per store per month. The dark store pays for itself — in direct store-level EBITDA terms — once it crosses approximately 1,000 orders per day, a threshold that 75% of Zepto's stores had crossed by May 2024, according to internal company metrics cited by multiple analyst reports that year.

The dark store is not a warehouse. It is a picking machine. Every design decision — shelf height, aisle width, SKU placement, packer position — is made with one question: how do we get from order placed to bag sealed in under three minutes?

Technology runs underneath all of this. Zepto's demand forecasting system updates SKU stock requirements in near-real time, factoring in weather, time of day, local events, and historical order patterns by neighbourhood. If Mumbai gets a monsoon forecast and searches for rain-related items spike — packaged chai, biscuits, instant noodles — the system flags replenishment orders to the nearest hub before demand peaks. The AI here is not a differentiator per se (Blinkit and Instamart have comparable systems), but the hyperlocal granularity of the training data — three years of order-level data by pin code — is genuinely hard to replicate from scratch.

Founders building retail or food businesses who want to understand how this type of demand forecasting works at a smaller scale should look at AI inventory forecasting for retailers — a category of SaaS tools that brings Zepto-style predictive replenishment to businesses doing ₹50 lakh to ₹10 crore in monthly revenue.

The Funding Flywheel: From $60M to $7 Billion in Four Years

Zepto's fundraising trajectory is almost as striking as its revenue growth. The company raised $60 million in December 2021 at a $225 million valuation — a Series A round led by YC Continuity. Six months later, in May 2022, it raised $200 million more and the valuation doubled to $900 million. By August 2023, Zepto had crossed $1 billion in valuation (officially a unicorn, though the founders were by then 21 years old). A $350 million round in November 2024 pushed the valuation to $5 billion. Then in October 2025, California Public Employees' Retirement System (CalPERS) — a US pension fund that rarely leads direct startup investments — anchored a $450 million round at a $7 billion valuation. Total capital raised: approximately $2.93 billion.

What drove investors at each stage? The answer changes by round. Early investors were betting on the model and the founders' execution speed. The 2022–2023 rounds were bet on proof that dark stores could reach EBITDA breakeven. The 2024–2025 rounds were bet on IPO optionality — a profitable Blinkit owned by Zomato had demonstrated that quick commerce could become a public-market story. CalPERS' October 2025 entry is the signal that institutional investors now believe quick commerce in India is structurally durable, not a venture-funded discount experiment.

By December 2025, Zepto had confidentially filed its Draft Red Herring Prospectus (DRHP) with SEBI, aiming to raise ₹11,000 crore (approximately $1.22 billion) in a public offering. As of May 2026, SEBI had granted approval, with the IPO window expected between July and September 2026. At a $7 billion valuation, the IPO would make Palicha — born in October 2001 — one of the youngest billionaires in Indian stock market history, if not the youngest, at just 24.

The capital structure underpinning Zepto's growth — and the role of foreign institutional money — is relevant context for founders exploring revenue-based financing for D2C brands, which offers an alternative funding path that does not require the rapid valuation-step-up pressure that Zepto's VC rounds created.

Zepto lost more money for every additional rupee of revenue it earned in FY25 than it did in FY24 — losses grew 177% while revenue grew 129%. That is the most important number in the story, and the one the IPO roadshow will need to explain.

The Unit Economics: What the Numbers Actually Say

Zepto's FY25 financials are simultaneously impressive and alarming. Revenue of ₹9,669 crore — up 129% from ₹4,224 crore in FY24 — is the headline. But the net loss of ₹3,367 crore (up 177% from ₹1,215 crore in FY24) grew faster than revenue. In FY25, Zepto's losses amounted to approximately 35% of its total turnover, up from 29% in FY24. Scale was supposed to bring cost efficiencies. In FY25, it did the opposite.

The explanation is structural rather than managerial. Quick commerce in India has three permanent cost centres that do not drop with scale in the way that, say, software infrastructure costs do. First, last-mile delivery: every order requires a rider, a route, and fuel — physical delivery has no digital scale effect. Second, dark store operations: rent, staff, refrigeration, and shrinkage are largely fixed per store, and adding stores multiplies these costs. Third, discounting and platform fees: Zepto has consistently offered below-cost pricing on many categories to defend market share against Blinkit and Instamart. The CAIT (Confederation of All India Traders) has alleged that Zepto and its competitors have collectively channelled ₹54,000 crore of foreign direct investment into subsidising customer acquisition through deep discounts — a practice that CAIT argues violates India's FDI policy, which prohibits inventory-based e-commerce from receiving foreign capital.

The contrarian read — and the one most serious investors are making — is that the unit economics at the individual dark store level are actually improving. By May 2024, 75% of Zepto's stores had reached full store-level EBITDA positivity. The consolidated loss figure is driven primarily by the cost of opening new stores (each new dark store takes six to twelve months to ramp to breakeven order volume) and by the corporate overhead of scaling a 1,000-location business. As store opening velocity slows — which it must, because the best locations in the top 10 cities are already taken — the consolidated economics should improve.

The inventory management discipline required to run a profitable dark store at scale shares structural similarities with the challenges facing any high-SKU retail operation. Our piece on inventory management for small Indian e-commerce sellers covers the reorder formulas and ABC analysis that Zepto's operations teams apply at the individual dark store level.

Competition: Blinkit, Instamart, and the Race No One Is Winning Yet

India's quick commerce market had a gross order value of approximately ₹64,000 crore in FY25 — more than double the prior year. Three players dominate: Blinkit (owned by Zomato, which rebranded from Eternal in 2025) with over 50% market share by September 2025, Zepto at approximately 22–29%, and Swiggy Instamart at 24%. Together, the three processed more than 4.15 million daily orders in March 2025 — Blinkit leading at 1.65–1.75 million, Zepto at 1.45–1.55 million, and Instamart at 1.05–1.15 million.

Blinkit — built by Albinder Dhindsa from Gurugram and now owned by Deepinder Goyal's Zomato — is the clearest benchmark for where Zepto needs to go. Blinkit reported an adjusted EBITDA profit of ₹37 crore in Q4 FY26, compared to a loss of ₹178 crore in the same quarter a year earlier. The swing is significant: it is the first major quick commerce player in India to demonstrate that the model can be profitable at scale, and it has done so at approximately 2,100 dark stores versus Zepto's 1,000–1,200. The implication for Zepto is that it needs to either add stores faster or wring margin improvements from its current store count — and the FY25 data suggests it has not yet achieved the latter.

The competition is intensifying. Reliance JioMart was processing close to 1.6 million daily orders by early 2025, and Amazon and Flipkart have both launched quick commerce pilots. The grocery market is large enough to support multiple players — kirana stores still command 92% of India's grocery retail as of 2024, per industry estimates — but the acquisition of urban quick commerce customers is a zero-sum game. Every order placed on Zepto is one not placed on Blinkit, and vice versa. This dynamic is what drives the deep discounting that makes the FY25 loss figures so hard to interpret.

The cloud kitchen and food delivery space is another direct adjacency that quick commerce platforms are now moving into. Zepto Cafe — launched in 2024 — is competing directly with cloud kitchen operators. Founders building in this space should study the catering and cloud kitchen platform model to understand where the margins are likely to settle once the deep-discount phase ends.

The Regulatory Minefield: FDI, CAIT, and the CCI

The most significant risk to Zepto's IPO timeline is not its loss widening — sophisticated public market investors understand loss-making growth companies. The risk is regulatory. India's FDI policy under Press Note 2 of 2018 permits 100% foreign investment via the automatic route for marketplace e-commerce models, but prohibits FDI in inventory-based e-commerce. The distinction matters: a marketplace connects buyers and sellers; an inventory-based model buys, holds, and sells stock directly.

CAIT, led by Secretary General Praveen Khandelwal (also a Lok Sabha MP), has filed formal complaints with the Enforcement Directorate alleging that Zepto, Blinkit, and Swiggy Instamart operate as de-facto inventory holders — company-owned dark stores stocking company-purchased goods — while classifying themselves as marketplaces to benefit from FDI. CAIT's white paper, published in November 2024, alleged that quick commerce platforms had collectively channelled ₹54,000 crore in FDI primarily into subsidising operations rather than the asset creation that FDI policy envisages. The same CAIT filing claimed that over two lakh kirana stores had closed since 2023 as quick commerce platforms captured urban grocery demand.

The Competition Commission of India (CCI) opened a preliminary inquiry into predatory pricing in the quick commerce sector in August 2025. Predatory pricing — selling below cost to eliminate competition — is a violation of the Competition Act, 2002. Quick commerce platforms argue that the discounts are promotional, temporary, and funded by venture capital, not cross-subsidised from another revenue line. The CCI has not yet reached a finding, and the inquiry is ongoing as of May 2026.

Zepto has attempted to address the FDI compliance question by structuring its holding company with an Indian-majority ownership at the entity that directly operates dark stores. Whether this structure satisfies the letter of Press Note 2 is a question that regulators, not the company, will answer. The risk is real: if the ED or Ministry of Commerce rules that Zepto's structure is non-compliant, it could trigger a requirement to restructure its cap table — an event that would almost certainly delay the IPO and affect its valuation.

What Founders Can Learn From the Zepto Playbook

The Zepto story is not a blueprint for replication — few founders will raise $2.93 billion and build 1,000 warehouses. But embedded in it are four operational principles that apply at any scale.

First: constraint is a feature, not a limitation. Zepto's decision to stock 8,000 SKUs instead of 80,000 looks like a weakness from the outside. It is the single most important operational decision the company made. Every supply chain, every demand forecasting model, every picker training programme became simpler and faster because the product catalogue was bounded. Founders who try to be everything to every customer typically succeed at nothing in particular. Pick your 8,000 SKUs.

Second: the location is the product. In quick commerce, the dark store's address is more important than its technology. Zepto's founders understood this early and raced to sign leases in prime residential micro-markets before Blinkit woke up to the model. In any business with a physical footprint — a dark store, a cloud kitchen, a diagnostics lab, a tier-2 city retail point — the location decision compounds over years. Make it first; make it right.

Third: unit economics at the micro level must work before you scale. Zepto knew the economics of a single Mumbai dark store before it expanded to Bengaluru. It knew what AOV it needed, what order density it needed, and what pick efficiency it needed. Founders who scale before the unit economics are proven are borrowing trouble. The losses become structural; the path to profitability becomes longer. Run your unit economics model at 10 stores before you plan 1,000.

Fourth: the regulatory environment is a real variable, not background noise. Every business operating in a space that touches FDI policy, platform-vs-inventory classification, competition law, or consumer protection needs to stress-test its structure against regulatory scenarios before scale makes the structure hard to change. Zepto is managing this challenge now, at IPO stage, under public scrutiny. Doing it at seed stage would have been simpler and cheaper.

For founders building businesses in the D2C logistics and shipping space that serves quick commerce suppliers and sellers, the D2C shipping aggregator model is one of the most capital-efficient ways to capture value in the quick commerce supply chain without owning dark store real estate.

Where Zepto Goes From Here

The IPO — expected in the July to September 2026 window after SEBI approval in May 2026 — is the next major milestone. At a $7 billion valuation and with plans to raise ₹11,000 crore, the offering will price Zepto against Blinkit (which trades as part of Zomato's market cap at approximately ₹2.3 lakh crore as of early 2026). Public market investors will pay close attention to three things: the trajectory of store-level EBITDA margins, the resolution of the CCI inquiry, and the path to consolidated profitability.

Beyond India, Zepto has no explicit international expansion plan, and none is needed. India's urban grocery market — currently 8% penetrated by organised retail — is large enough to sustain three to four profitable quick commerce players at 10x Zepto's current scale. The question is not whether the market is big enough. The question is whether Zepto can convert its first-mover infrastructure advantage into lasting margin before a better-capitalised competitor (read: Reliance JioMart) narrows the operational gap.

Palicha, now 24, told Business Standard in April 2025 that he sees Zepto as "building an internet supermarket chain" — not just a delivery app, but a vertically integrated grocery retailer with the selection, quality, and price competitiveness of a large organised grocery chain combined with the speed of a neighbourhood kirana. If he achieves that, and the unit economics prove out, the FY25 losses will look like exactly what he says they are: the price of building category-defining infrastructure. If he doesn't, they will look like what the sceptics say: the most expensive grocery experiment in Indian startup history.

Last updated: May 2026