Grocery integration
into Same Day network

Two apps for one customer — the classic conversion killer

Amazon's grocery in India lived in a walled garden: a separate "app-in-app" experience that required customers to navigate away from their primary shopping session. If you were buying a phone charger and wanted to add milk to the same order, you had to mentally switch contexts — and most customers didn't.

The downstream effects were expensive. Customers placing separate orders for grocery and non-grocery items meant multiple delivery riders dispatched to the same address. Multiple shipments. Multiple delivery attempts. Higher cost per order. Higher carbon footprint. And still — lower customer satisfaction because waiting for two deliveries is worse than one.

The fragmentation wasn't just a UX problem. It was a unit economics problem.

Integrate vs. separate surface — why this fight mattered

There was a strong internal push to build a standalone quick-commerce grocery surface. The argument was reasonable: cleaner scope, faster delivery, less cross-team dependency, a purpose-built UX for the grocery persona.

I opposed it — and built the case for deep integration into the existing Same Day app. My reasoning:

• Basket abandonment is invisible on separate surfaces. You can't see the customer who decided not to add milk because they'd have to open another app. The lost basket never shows up in your metrics — it just disappears.
• Single order, single shipment is the value proposition. Customers don't want two delivery windows. They want one. Any architecture that requires two is already half-broken.
• Separate surfaces fragment inventory visibility. Real-time stock for grocery and non-grocery items needs to be visible in a single fulfillment view — otherwise promising customers items you don't have becomes routine.

End-to-end integration across the fulfilment stack

This wasn't a front-end redesign. It required rebuilding how grocery inventory, orders, and deliveries flowed through Amazon's Same Day infrastructure:

• Unified shopping experience: Grocery catalogue surfaced inline within the main Same Day app, with real-time stock and delivery promise displayed at item level — no separate entry point required.
• Single-order, single-shipment fulfilment model: Technical integration across WMS, last-mile, and logistics coordination to merge grocery and non-grocery items into one shipment per order wherever possible.
• Vendor onboarding at scale: Structured onboarding pipeline enabling 1,000+ grocery vendors — with standardised cataloguing, inventory feeds, and SLA requirements baked in from day one.
• SKU eligibility engine: Automated framework determining which of the 25K+ SKUs could be included in same-day fulfilment based on shelf life, handling requirements, and DC proximity.
• Real-time inventory visibility: Cross-system stock visibility layer ensuring customers were never promised items the FC couldn't fulfil within the same-day window.

The numbers

• +18% conversion rate — customers adding grocery to existing Same Day baskets instead of abandoning or ordering separately
• 25,000+ SKUs onboarded into the integrated same-day network
• Double-digit reduction in total customer shipments — fewer deliveries per order, lower cost, lower environmental impact
• Maintained delivery SLA — same-day promise held despite adding grocery complexity into the FC picking and dispatch flow

Other projects from this era