Sub-20-minute delivery and 3x rider productivity from dark stores

This q-commerce brand runs a network of dark stores promising grocery delivery in minutes. The promise is also the hardest part of the operation: orders arrive in bursts, riders are a mix of employees and gig workers, and every minute between packing and doorstep counts. Manual assignment left riders idle between trips, orders weren’t batched efficiently, and missed promise times led to refunds and churn.

Geofleet connected to the order and dark-store systems to assign and batch deliveries in real time. As soon as an order is packed, the system routes it to the best available rider — combining nearby orders into efficient batches based on location, capacity, and prep status. Customers receive a live ETA and tracking link, and a surge mode adjusts promise times and rebalances riders when demand spikes.

Average delivery time dropped below 20 minutes, and orders delivered per rider per hour roughly tripled as idle time and backtracking fell away. On-time rate reached 98%, refunds for late orders fell sharply, and the brand could open new dark-store zones confident the delivery layer would scale with them.

Tripling rider productivity also reshaped unit economics: the cost of fulfilling each order fell because the same rider hours produced far more deliveries, which is decisive in a category where every order carries a thin margin.

In q-commerce, speed is the product. A customer chooses a rapid-grocery brand precisely because of the sub-20-minute promise, so missing it doesn’t just cost a refund — it erodes the only real differentiator the brand has. Making that promise reliable at 98% on-time protected both retention and the brand story.

Productivity is the other half of the equation. Because each delivery earns so little, a rapid-grocery operation only works if rider hours are used efficiently. Tripling orders per rider per hour is what turns a fast service into a viable business.

Dispatch stopped being a manual scramble during bursts. Orders flow to the best rider automatically the instant they’re packed, so the operations team manages exceptions and surges rather than assigning every order by hand.

For riders — a mix of employees and gig workers — the day became less about waiting and guessing and more about a steady, sensible flow of batched stops. And with surge mode handling demand spikes, the team could expand into new dark-store zones knowing the delivery layer would hold up.