Zepto: How Two 17-Year-Olds Built India's Largest Seller Of Fruits and Vegetables

Zepto’s journey began with a first-principles obsession: imagine the most extreme positive customer experience and work backward to make it real. Founded during the pandemic by two young co-founders who started by delivering groceries to 30 local customers via WhatsApp, the company pivoted from a shop-delivery model to company-owned mini dark stores after realizing that controlling fulfillment was the only way to truly satisfy customers. This customer-led innovation—not competitive copying—unlocked 3–4x volume and validated their radical 10-minute delivery promise, despite skepticism from industry outsiders. From there, Zepto scaled into India’s largest fruit and vegetable supply chain, processing millions of daily deliveries, and is now building an ad business while aiming to become the country’s urban grocery infrastructure. The core lesson is that deep customer conversations, relentless operational focus, and youthful naivety—rather than grand strategy—turned a simple neighborhood experiment into a billion-dollar business.

• ▶ 0:00 The founding philosophy was to remove constraints and think from first principles about the most extreme positive customer experience possible, then work backward to make it feasible.
• ▶ 1:25 At 17, the co-founders pivoted during the pandemic in Mumbai, starting by delivering groceries via a WhatsApp group chat, which evolved into their first app, Kiranakart.
• ▶ 3:52 Their advice to founders: don't fully drop out; take a year off, work to reach product-market fit (seen around 8–9 months), and only make a major commitment when the business is proven — roughly 10,000 orders per day.
• ▶ 4:45 Early customer traction ("working customers are liking it") became the signal that convinced the founders to pursue the idea.
• ▶ 4:47 India's grocery delivery market already had $2–3 billion incumbents, making the space look saturated long before Zepto entered.
• ▶ 5:04 The founders' key challenge was justifying entry against well-funded, scaled competitors—where most people would conclude "it's too late, I missed it."
• ▶ 5:55 The founders built conviction through direct customer conversations at doorsteps, asking why customers used them, why they avoided competitors, and why they didn't just shop themselves—revealing that end customers were deeply unsatisfied with existing options.
• ▶ 6:24 Early feedback showed the first product version lacked selection, had high pricing, and offered uncontrollable delivery times—prompting the founders to ask what customers truly cared about.
• ▶ 6:44 Customer conversations surfaced four core priorities—speed, quality, selection, and price—which led to experimenting with mini dark stores and achieving genuine product-market fit.
• ▶ 7:02 Zepto was a pivot from an earlier company called Coronicart, developed during Y Combinator.
• ▶ 7:30 The founders initially solved a simple neighborhood problem for 30 aunties using a WhatsApp group chat, taking orders and delivering from local stores.
• ▶ 8:07 In early 2021, they pivoted to mini warehouses after realizing the shop-delivery model gave no control over customer experience or delivery times, creating their first dark store in KB's apartment.
• ▶ 8:15 The first dark store was the co-founder's apartment, serving as the starting point for Zepto's delivery model.
• ▶ 8:20 The controlled, small-scale setup produced a phenomenal customer response, validating their approach and encouraging scaling.
• ▶ 8:30 They then built a "mini cardboard version of a warehouse" in Juu as an intermediate step before launching a proper warehouse.
• ▶ 8:40 The pivot was triggered by the realization that they needed to control the customer experience, since incremental improvements only produced marginal gains.
• ▶ 9:01 Data showed dark stores generated 3–4x the volume of mom-and-pop delivery, a 10x demand unlock that proved the new model's potential.
• ▶ 9:13 This evidence drove the decision to pivot to company-owned mini warehouses (dark stores) to own the fulfillment experience directly.
• ▶ 9:14 The mini warehouse and doorstep delivery model was novel at the time and came from internal thinking, not borrowed from others.
• ▶ 9:26 The founders did not copy competitors; they focused on customer dissatisfaction and what could be done operationally to raise satisfaction.
• ▶ 9:44 Despite outsiders saying it wasn't the right way, customers validated the model, showing innovation was customer-led rather than industry-driven.
• ▶ 9:51 Skeptics said building dark stores was wrong because many businesses built models "backwards from a supply chain or from financial goals," forcing a viable model onto customers.
• ▶ 10:18 The conventional model treats customers as passive receivers: "this is the model that's viable and this is what I'm going to give to the customer... this is what the customer is going to get."
• ▶ 10:24 Zepto inverted that logic by asking what extreme the customer wants, which led to the radical 10-minute delivery promise — still "crazy" and mind-blowing to the interviewer.
• ▶ 10:38 Customer delight is the first-principles goal: imagine the most extreme positive customer experience without constraints, then work backwards to deliver it at scale — this drove Zepto's 10-minute delivery model and created a virtuous loop of higher volume and lower costs.
• ▶ 12:54 The founder pushes back on the "iterative genius" narrative: they were among the worst-performing companies in their YC batch, and their progress came from youth and naivety, not a grand strategy — not knowing how hard the problem was became a huge advantage.
• ▶ 14:22 Zepto embodied the YC principle of "make 100 users love you": focusing obsessively on just 30–40 early customers revealed that the original "Corona" model was wrong and showed them what the right experience (Zepto) should be.
• ▶ 15:38 Zepto's core identity is a supply chain and grocery company ("atta-dal" staples), not just a software app, as reinforced by customers comparing it to a local kirana store.
• ▶ 16:14 Massive operational depth exists beyond the app: dark store design, replenishment algorithms, trucking, 200,000+ employees, and industrial-grade automation for suppliers like Unilever and Coca-Cola.
• ▶ 17:53 Zepto now runs India's largest fruit and vegetable supply chain, sourcing millions of units weekly from farmers directly, and processes millions of deliveries per day with topline in the billions of dollars.
• ▶ 19:26 Zepto's new advertising business, where major brands bid on search keywords, is already generating "hundreds of millions of ARR in ad revenue," and was "very small or inconsequential" just two years ago.
• ▶ 20:20 The 10-20 year vision is to build "urban grocery infrastructure for India," positioning Zepto as an "organizing force" in the supply chain, with near-term plans to become a large grocer in the top 40-50 cities and potentially create "India's homegrown version of Amazon."
• ▶ 22:10 Beyond commerce, the founders emphasize ground-level employment and helping "build a small part of India's infrastructure," while also serving as a launchpad for new consumer brands.

• ▶ 22:44 The conversation shifts to AI, noting Zepto was founded before the AI boom but is now at the cutting edge of AI implementation.

• ▶ 23:01 Zepto has used AI to transform its internal operations, a point referenced from an earlier morning conversation.

• ▶ 23:07 The interviewer asks the founders to elaborate on exactly how Zepto is using AI and how it has changed the company.

• ▶ 23:20 Previously, manual forecasting of the supply chain took a team days to complete.
• ▶ 23:31 Zepto built an ML algorithm that forecasts millions of units per day with no humans in the loop.
• ▶ 23:39 ML-driven forecasting made the supply chain more agile, enabling faster dispatch, higher throughput in the same space, and quicker forecasts.
• ▶ 23:51 AI/generative AI on the consumer side, especially in the ads business, drove a big bump in ads revenue over the past two years.
• ▶ 24:05 Built search-platform tools that let brands generate keywords and a model that predicts which keyword yields the best return on ad spend (ROAS).
• ▶ 24:18 Advertising growth was massive because brands could objectively see marketing dollars used optimally via ROAS metrics, leading to a significant revenue increase.
• ▶ 24:38 The company is doing "a lot more with the same headcount," with a small engineering team that is remarkably productive.
• ▶ 25:04 They've cut almost all software spending to zero and drastically reduced managed-service costs.
• ▶ 25:14 They removed manual inefficiencies through automation, replacing outsourced tools with in-house automated tools, yielding huge cost savings.
• ▶ 25:41 Zepto's software team has roughly 500 engineers, plus about 150 in data science, analytics, product, and design.
• ▶ 25:51 Zepto is "very much hiring" and actively seeking top talent, with an open invitation for interested listeners to reach out.
• ▶ 26:01 Zepto still considers itself an early-stage startup, not a "big old slow company," and invites those excited by fast-paced culture to join.
• ▶ 26:12 The interviewer signals time is running out and introduces the last question of the conversation.
• ▶ 26:18 The final question centers on how the founder mastered an incredible number of skills at a very young age.
• ▶ 26:22 The question begins to pivot toward the founder's origin story ("you started this…") before the transcript cuts off.
• ▶ 26:49 Maintain humility and perspective: Kaivalya credits Y Combinator for grounding him and views Zepto's work as "peanuts" compared to companies like Uber or Airbnb, keeping him motivated on a long journey.
• ▶ 27:09 Surround yourself with smarter people: The core strategy was hiring senior, experienced leaders (CFO, COO, CTO, etc.) who were also hungry and believed in the vision, forcing him to "pull up his boots" and learn from them.
• ▶ 28:24 Learn shamelessly by asking basic questions: Kaivalya still asks simple questions to finance and ops teams, admitting he "looked very stupid" internally but aggregating knowledge through shameless questioning. Final advice: "Surround yourself with people that are smarter than you and learn from them shamelessly."

• ▶ 0:00 First Principles, Origin Story, and Early PMF (0:00 - 4:47) - - First-principles thinking, how Zepto started, and early signs of product-market fit.
• ▶ 4:47 Pivoting to Dark Stores and Customer Obsession (4:47 - 10:41) - - Competing with incumbents, the original idea, and validating dark stores by putting customers first.
• ▶ 10:41 Iterative Building and YC Discipline (10:41 - 15:08) - - Customer delight from first principles, scrappy iteration, and YC's focus on only team and customers.
• ▶ 15:08 Zepto's Real Identity: Supply Chain and Scale (15:08 - 19:15) - - Zepto as a kirana-style daily-groceries company with deep operations, a huge fruit/vegetable supply chain, and massive scale.
• ▶ 19:15 Long-Term Vision and Market Opportunity (19:15 - 22:44) - - Efficiency and advertising revenue, the long-term grocery-infrastructure vision, and quick-commerce as a launchpad for consumer brands and jobs.
• ▶ 22:44 AI, Hiring, and Founder Advice (22:44 - 28:58) - - AI across supply chain and ads, cost efficiency, software hiring, and closing advice on humility and learning from smarter people.