Every October, India's online retail machinery is asked to do in a few weeks what it normally does in a quarter. For most of the last decade, that strain fell on the marketplaces and the express networks that served them. This year, for the first time, a large share of it is falling on a channel built to deliver a bottle of milk in ten minutes.

By August, India's quick commerce platforms were already handling around 9 Mn orders a day across more than 7,000 dark stores, up 47% on the year by Datum Intelligence's count. Then the festive season opened on top of that base. In the first days of the sales, the quick commerce arms of Amazon and Flipkart ran at four to five times their normal daily GMV, according to Redseer, as both marketplaces began fulfilling festive deals through their own ten-minute networks.

Every one of those orders ends the same way: a rider, a vehicle, and a doorstep, on a promise measured in minutes. Whether that rider is still moving at the end of a 14-hour shift, with the Diwali peak still ahead, is the question this season will actually be judged on.

A Different Kind Of Growth

The numbers behind the season explain why. Redseer expects online retail to grow 25% this festive period, the strongest in five years, and projects quick commerce to grow 110%-120% against 16%-18% for the rest of e-commerce, taking its share of festive GMV from 11% to roughly 18%. Datum reads it more conservatively, noting that much of the growth is grocery shifting from scheduled delivery to instant, but even on that view the channel's share of the season rises from 12% to 16%. The precedent was set last Diwali, when Unicommerce's platform data showed conventional e-commerce orders rising 24% while quick commerce orders grew 120%.

Datum makes a sharper observation about the season as a whole: all of the growth is coming from more orders, not bigger ones. Twenty million more shoppers are buying, and buying more often, while the average basket shrinks. A season that grows on discount depth strains the balance sheet. A season that grows on volume strains the last mile.

Quick commerce is where those two pressures meet. Smaller baskets, more trips, and a delivery window that leaves no room to batch.

A Crowded Map

The competitive map has been redrawn around this. Flipkart Minutes, barely two years old, crossed 1,200 micro-fulfilment centres this autumn, is adding more than 100 a month, and by CLSA's count has already overtaken Swiggy Instamart on dark stores in the top 10 cities. Amazon Now, BB Now and JioMart have joined a field of at least seven multi-category players, most of them fighting for the same pincodes.

Eternal remains the benchmark, and it is the benchmark for a reason that matters here. Blinkit handled 917 Mn orders in FY26 and has since grown past 2,400 dark stores. But the number that tells you how the company thinks about the last mile is a different one: by the end of September, more than 2 Lakh delivery partners across Zomato and Blinkit were riding electric two-wheelers, up from just over 1 Lakh at the end of March. For the market leader, electric delivery is no longer an ESG footnote. It is an operating policy.

What The Rider Is Actually Being Asked To Do

The industry's response to the surge is visible in hiring. Adecco expects festive gig demand across ecommerce and quick commerce to rise 15%-20%, to as many as 2.7 lakh workers, with quick commerce accounting for 40%-45% of that requirement, up from roughly a third last year.

The hiring numbers hide an uncomfortable detail. The per-order rate itself has barely moved. Across the platforms our riders serve, base pay still sits around ₹20-30 an order; the festive uplift comes entirely from incentives for attendance and volume. In other words, the platforms are not paying more per delivery. They are paying more for riders who stay on the road and keep delivering.

Which means a rider's festive earnings now depend almost entirely on two things: how many orders they can complete and how little each one costs them to complete.

That arithmetic is what makes the electric two-wheeler the default quick commerce vehicle. A rider on a petrol scooter watches every additional kilometer add to cost; the surge erodes margin even as it lifts volume. A rider on a rented EV, paying a flat weekly fee for the vehicle, battery, charging, and service, sees the opposite. More orders mean greater earning potential at the same running cost. But that advantage only holds while the vehicle is moving.

Capacity Is Not Uptime

The harder problem is not getting riders onto electric vehicles. It is keeping them on the road. A ten-minute promise has no slack in it. A rider who cannot find a charger, or whose vehicle develops a fault mid-shift, earns nothing while the delivery window closes and the dark store's throughput drops. Every new store, every onboarded rider, every incentive that pushes for one more order, is a bet that the vehicle underneath will hold. For a fleet operator, the festive season is less a demand problem than an availability  problem.

Solving it means treating charging, battery health, and service as part of the dark store's operations rather than as the rider's private burden. A charger belongs where the loading happens so that topping up a battery costs minutes rather than a detour across the city. A battery pack should be reporting its own condition continuously so that a weak cell is caught at the store and not on a flyover with a bag of groceries going warm. And the rider's cost should be a number they know on Monday, not one that moves with every kilometer.

Our own experience this season has been a test of exactly that design. The fleet across Bengaluru and Gurugram has grown about 67% since July, and in the week of October 12, 293 vehicles went out in two days as around 300 new riders joined. What kept that ramp from turning into a maintenance queue was the infrastructure already in place: 6 kW DC chargers at the stores themselves, about ten vehicles to a gun; battery packs with cells immersed in dielectric fluid, each reporting to a monitoring platform that flags faults before they stop a vehicle; and a flat weekly rental between 1500-1700  that covers vehicle, battery, charging and service. None of that is visible to the customer waiting for a delivery, but efficiently influences the earning and longevity potential of a rider. 

What The Season Is Really Testing

The festive surge is an annual stress test for the last mile. This year the load is a quick commerce base projected to nearly double in a single season, and the response so far has been capacity: more stores, more riders. Capacity is what gets announced but a layer deeper.  Its uptime is what gets measured through the operations. The platforms that come out ahead will be those whose vehicles did not stop, whose riders earned enough to stay on, and whose unit economics makes the most sense. 

Eternal has shown what a 2 lakh-rider electric fleet looks like in production. Flipkart, with a 20,000-plus electric fleet and a pledge to go fully electric by 2030, is moving fast. The question for the rest of the market is no longer whether electric fleets can carry the peak. It is whether the charging networks, battery intelligence, and service operations underneath them were built for the peak at all.