Y CombinatorZepto: How Two 17-Year-Olds Built India's Largest Seller Of Fruits and Vegetables
CHAPTERS
- 0:00 – 0:17
First-principles customer experience: start extreme, then work backward
Aadit opens with Zepto’s core product philosophy: imagine the best possible customer experience without constraints, then engineer backward to make it real. This framing becomes the throughline for later decisions like dark stores and 10‑minute delivery.
- •Use first principles to define an “unbelievable” customer experience
- •Work backward from the ideal rather than forward from constraints
- •Customer delight is treated as the root of business value
- •Sets the conceptual foundation for Zepto’s model choices
- 0:17 – 2:25
From teenagers in Mumbai to a COVID WhatsApp grocery service
Aadit explains how he and his cofounder Kewal started building projects as teens, then returned to Mumbai due to COVID. They began by serving neighborhood grocery needs via a simple WhatsApp group, which evolved into an app called Kirana Cart.
- •Founders began at 17; originally building for fun, not “to start a company”
- •COVID drove them back to Mumbai and created urgent grocery access problems
- •Started with hyper-local deliveries for neighbors using WhatsApp
- •The early product became an app named Kirana Cart and later connected to YC
- 2:25 – 4:47
Stanford vs startup: delaying the leap until early PMF signals
Jared recounts Aadit’s decision point: go to Stanford or pursue the startup. Aadit describes a pragmatic approach—take a year, look for retention and compounding, and commit only after clear proof the business was real.
- •Decision framed as a risk-management problem, not a heroic leap
- •Took time to validate retention and early product-market fit
- •Committed after reaching ~10,000 orders/day and meaningful run rate
- •Investor interest/term sheet plus customer love helped confirm the choice
- 4:47 – 7:03
Entering a crowded market by listening at the doorstep
Despite existing large grocery-delivery players in India, Aadit explains they found customers broadly dissatisfied. By doing deliveries themselves and talking to users in person, they learned what mattered most and why incumbents weren’t meeting needs.
- •Crowded market didn’t deter them because customer pain was still high
- •Founders did deliveries and conducted organic “doorstep interviews”
- •Identified key customer priorities: speed, quality, selection, price
- •Early Kirana Cart experience still lacked strong PMF and exposed gaps
- 7:03 – 9:54
The pivot to dark stores: gaining control of the customer experience
Aadit details the shift from delivering out of existing mom-and-pop shops to holding inventory in mini warehouses. Starting with stock in a cofounder’s apartment, they saw a step-change in demand once they controlled selection and delivery reliability.
- •Original model lacked control over delivery times and experience
- •First “dark store” was inventory stored in the cofounder’s apartment
- •Progressed from apartment → DIY warehouse → proper warehouse launch
- •Dark-store neighborhoods drove 3–4x volume, proving stronger demand
- 9:54 – 12:04
Why 10-minute delivery: customer-first beats constraint-first planning
They discuss skepticism about dark stores and rapid delivery, and Aadit contrasts two mental models: building backward from unit economics vs. starting from customer extremes. He argues nailing the experience can unlock unexpected operational advantages through higher demand and throughput.
- •Critics optimized for viability first and gave customers “what’s possible”
- •Zepto started from what customers would love most (10-minute delivery)
- •High demand density drives throughput that can lower unit costs
- •Customer delight is presented as prerequisite to profitability
- 12:04 – 15:07
Iterating fast and staying noise-free: YC’s ‘100 customers’ in practice
Jared highlights Zepto’s experimentation culture—constant tweaks, field learning, and humility about failures. Aadit attributes speed of learning to being young, focused on a small set of users, and insulated from startup “noise.”
- •Rapid experimentation; many failures and course corrections
- •Naivety helped them act without overthinking historical baggage
- •COVID environment reduced distractions to “just us and the customer”
- •YC principle: make a small number of users love the product
- 15:07 – 17:39
Zepto as a logistics, retail, and supply-chain company (not just an app)
Aadit reframes Zepto as a day-to-day grocery business powered by deep operational infrastructure. He describes the many behind-the-scenes systems—from dark-store design to replenishment, trucking, workforce management, and automation—that make 10-minute delivery possible.
- •Zepto sells everyday staples; customers don’t perceive it as a “tech company”
- •Operational depth: store layout, selection, replenishment, task systems
- •Scale requires managing a massive workforce across roles
- •Back-end automation and robotics reduce cost and improve speed/selection
- 17:39 – 20:06
Fruits & vegetables engine and Zepto’s scale today
The conversation moves to Zepto’s produce supply chain and the scale it has reached. Aadit describes sourcing directly from farmers and serving millions of units, plus the broader platform scale in users, orders, and revenue.
- •Built one of India’s largest fruits-and-vegetables supply chains
- •Sources large volumes directly from farming regions across India
- •Millions of deliveries per day; large monthly transacting user base
- •Business includes additional levers like a growing in-app ads revenue stream
- 20:06 – 22:44
Long-term vision: urban grocery infrastructure and hyperlocal e-commerce platform
Aadit outlines Zepto’s ambition to become a foundational organizing layer for India’s grocery and logistics infrastructure across major cities. He connects grocery dominance to a broader e-commerce platform opportunity and to enabling new consumer brands and employment growth.
- •Goal: build urban grocery platform/infrastructure for India
- •Near-term: become a leading grocer across ~40–50 cities
- •Long-term: India-specific hyperlocal e-commerce model at large scale
- •Platform can enable new consumer brands and create significant employment
- 22:44 – 25:35
How Zepto uses AI: forecasting, ad tech, and internal automation
Aadit breaks AI usage into supply chain and customer-facing applications. He highlights automated demand forecasting without humans-in-the-loop, AI-optimized advertising tools for brands, and broad internal productivity gains that reduce SaaS and manual work.
- •Supply chain: ML forecasting improves speed, accuracy, and throughput
- •Consumer/ads: AI helps generate keywords and predict ROAS for brands
- •Internal: automation boosts output per employee and reduces headcount pressure
- •Reduced third-party software spend and managed services via in-house tooling
- 25:35 – 26:12
Engineering organization and hiring
Zepto’s technical org size and composition are summarized, along with an explicit hiring pitch. Aadit positions the company as still “early-stage” in mindset despite its scale.
- •~500 engineers plus data science/analytics/product/design talent
- •Emphasis on startup pace and building across software + hardware
- •Actively hiring for top talent
- •Frames Zepto as still early in its journey
- 26:12 – 28:57
Leveling up as a founder: learn shamelessly and hire people smarter than you
Aadit explains how he grew into operating a complex, large-scale business at a young age. His main recipe is surrounding himself with experienced, hungry leaders and asking basic questions without ego.
- •Personal growth driven by proximity to exceptional senior leaders
- •Early management hires (CFO/COO/CTO/etc.) became key teachers
- •Asking “stupid” questions is encouraged to accelerate learning
- •Core principle: surround yourself with smarter people and learn constantly
