India’s ten-minute delivery platforms run on daily replenishment, dark store inventory, and hyperlocal demand. Most brand forecasting still runs on a weekly rhythm built for modern trade. The mismatch is costing sales and ad spend.
For most of the last decade, Indian FMCG and personal care brands planned demand much the same way. A monthly sales and operations plan, refreshed weekly. Forecasts at the region and channel level. Distributors and marketplace warehouses held weeks of cover, which smoothed out most errors.
Quick commerce removed that cushion. Blinkit, Swiggy Instamart, Zepto, and the newer rapid services from Flipkart and Amazon fulfill orders from dark stores that each carry a limited assortment and hold very little stock per SKU. Replenishment happens daily, sometimes more often. Demand is shaped by the pin code, the weather, the day of the week, and whatever is happening locally. A weekly forecast averaged across a city cannot see any of that.
The unit of planning has changed
In marketplace and modern trade planning, the unit is roughly SKU by warehouse or SKU by region, per week. In quick commerce, the unit that decides whether you make a sale is SKU by dark store, per day. That shift has three effects.
Availability fragments. A SKU can be in stock across a city overall and still be missing from the dark stores that serve your highest demand neighborhoods. City-level averages hide this completely.
Demand spikes are local and short. A rainy evening, a cricket final, or a festival in one part of a city moves demand in a handful of dark stores for a few hours. By the time a weekly review sees it, it is over.
Lost sales go unrecorded. When a shopper opens the app and your SKU is unavailable, they buy a substitute. Your sales history shows lower demand, not a stockout. Next week’s forecast learns the wrong lesson.
Where the damage shows up
The first place most brands notice it is ad spend. Quick commerce platforms have built their own retail media businesses, and brands are quickly moving budget into them. But one team usually runs ads, and another manages availability. Spend keeps flowing to SKUs that are unavailable across large parts of the city, sending shoppers to a listing they cannot buy, or straight to a competitor.
The second is replenishment. Purchase orders from quick commerce platforms are frequent and variable. Brands that forecast weekly respond late, fill rates slip, and platform algorithms start favoring suppliers who keep dark stores stocked.
The third is promotions. A discount that sells out in the first few hours in most dark stores looks like a weak promotion in the data, when it was actually a supply failure.
What forecasting has to become
Catching up does not mean scrapping your planning process. It means adding a faster layer beneath it.
- Forecast at the grain where sales are lost. Dark store, or at least dark store cluster, per day, for the SKUs that drive your quick commerce revenue.
- Correct for hidden demand. Treat stockout periods as missing data rather than low demand, so the model does not learn from empty shelves.
- Bring in local signals. Platform search trends, weather, local events, competitor availability, and your own ad activity all move quick commerce demand faster than they move other channels.
- Link ads to availability. The simplest high-return change is a rule, or better, an agent, that pauses or shifts spend when a SKU is unavailable in the areas a campaign targets.
- Shorten the loop with the platform. Share forecasts with the platform’s category and supply teams in a format they can act on, and respond to purchase orders the same day.
Getting the media side right first
For many brands, the fastest win is on the ad side, because the waste is visible and the fix sits within the marketing team’s control. If you are still setting up on the largest platform, a practical guide to Blinkit ads covers the campaign types and structure you need to get right before layering availability-aware pacing on top.
Once media is tied to availability, the forecasting work has a clear business case. Every rupee of spend that no longer lands on an empty shelf is a gain you can measure, and it funds the harder work of forecasting at dark store level.
The weekly cycle is not coming back
For many urban categories, quick commerce is no longer a side channel. It is where impulse buys, top-ups and a growing share of planned purchases happen. The brands that treat it as a daily, hyperlocal business will set the standard for availability on these platforms. The rest will keep planning for a week their shoppers no longer live in.
