Kirana Stores Don't Need Saving. They Need Demand.

Before sunrise, a kirana owner stands in front of a half-empty shelf and makes the decision that shapes the week: how much to buy. Tomatoes or not? Another carton of the oil that moved well last month? There is no dashboard. There is memory, instinct and a long relationship with the people on the street. Buy too little and a customer walks out. Buy too much and the money sits on the shelf until it spoils or the price comes down.

Every purchase commits cash before demand is known. The shopkeeper has something valuable: years of experience, a close relationship with the neighbourhood and an understanding of local buying habits. What is missing is a reliable, affordable way to turn that knowledge into a demand signal before buying stock. That is the problem digital enablement should solve.

The kirana is not a legacy format. It is the system.

Roughly 12 million grocery retail outlets, most of them small shops, serve India, according to McKinsey’s 2022 report on Indian grocery retail. A 2022 CB Insights analysis puts the kirana share of India’s FMCG sales, the packaged everyday goods on a grocery shelf, at just over 90%, a share it says grew during the pandemic.

They last because of what no warehouse can copy: proximity, trust, credit on a handshake, the freedom to break a pack, and a shopkeeper who knows which family has stopped eating onions for the festival. Those are not features an app adds later. They are why the format has outlasted supermarkets, big-box chains and every wave of e-commerce.

The threat is not technology. It is the gap in tools.

Competitors are investing heavily. In 2024, more than two-thirds of India’s online grocery orders went through quick-commerce platforms, according to Bain & Company’s How India Shops Online 2025, written with Flipkart. That shows how much capital is chasing the category. It does not show which model works best for neighbourhood retail.

Those platforms can invest in forecasting systems and in stock held close to the customer. A kirana owner has none of that, and nobody should expect the owner to buy it.

Calling kiranas backward is the wrong diagnosis. They are under-tooled.

What digital enablement usually means, and where it falls short

Each of these tools solves a real problem. But recording a sale, enabling a payment, expanding delivery reach and simplifying restocking are not the same as knowing what a household intends to buy before the store commits its money. The missing layer is the connection between household demand and the kirana’s purchasing decision.

The missing layer is demand

Fast delivery depends on prediction: stock placed close to the customer, bought ahead of any order, in the hope that the forecast was right. A better forecast improves the estimate. A planned order gives the store a demand signal from the household before it procures the goods.

Here is the chain we are testing:

1. The household plansIt states what it needs for the coming days.
2. The store receives the orderA neighbourhood kirana gets that planned order before it buys stock.
3. The store sources to the orderIt buys what the orders require, from the mandi or its distributor.

The difference is that the model captures demand before procurement. It does not try to predict demand more accurately. What we hope follows, and what the pilot has to test, is that the store buys less on speculation and makes better use of its working capital. A planned order is a demand signal, not a guarantee: availability, substitutions, cancellations and last-minute changes still have to be managed.

How a planned order works today

How does a household send its list?

In the ShopConnect app it builds a list, from a week's meal plan or item by item, and places it with a neighbourhood store for a delivery day. Orders for a day close at 11 pm the night before, or at the store's own cutoff.

How does the order reach the store?

The owner gets an SMS and an app notification, and confirms the order in the ShopConnect store app. Once orders close, the app totals the next day's orders item by item, so the owner knows what to buy before sourcing.

What if an item is unavailable?

The store marks what it could not pack, and the household pays only for what arrives: a cash order's total drops straight away, and a missing item on an order paid online is refunded once the household reports it.

Can the store substitute an item?

Not yet. There is no step for a store to propose a replacement and the household to approve it, so an item the store cannot get is marked as short. How substitutions should work is one of the questions the pilot is testing.

How is it delivered and paid for?

The store delivers in its own delivery slot, or the household collects from the shop. The household can cancel until orders close. The store collects cash or UPI at the door; for an order paid online, the store's share is paid out once the return window closes.

Run a kirana or plan your household’s groceries? See how to take part.

It helps to separate three things. Demand capture makes it easier to order from a local kirana. Demand fulfilment helps the store source and fulfil. Demand creation, new purchases that would not otherwise have happened in the neighbourhood, is a hypothesis we have not yet measured. Together we think of this as demand-led neighbourhood commerce.

What good enablement looks like

These are the tests we set for ourselves, not claims of what is already proven.

• It should run on the phone the owner already has.
• It should work in the owner's language and in the tools already in daily use, such as WhatsApp.
• It should leave the owner in charge of the shop, the customer relationship and the prices.
• It should bring demand, not only tools.
• It should not ask the owner to pay for software before it has brought an order.

Testing the idea in Indiranagar

In Indiranagar, Bengaluru, we are testing a simple proposition: what changes when a neighbourhood kirana receives a household’s grocery order before sourcing the goods? In the model we are testing, households plan their grocery lists, orders reach a neighbourhood store, and the store sources from the mandi after the order arrives. The aim is to connect purchasing decisions more closely to what households have planned to buy, rather than relying entirely on forecasts or instinct. The model is designed to reduce speculative purchasing.

We are still in the early stages and do not yet have published results. The model also raises operational questions: how to manage many small orders and settlements, fulfil each day’s commitments, and keep sourcing reliable and economical. Inventory-led models carry working capital and write-off risk instead. Neither is simple. They are different.

What we plan to track

• DemandIncremental orders against a baseline, separating new demand from orders shifted from another channel.
• ConversionPlanned orders that become completed purchases, with cancellations and substitutions.
• Store economicsGross margin after procurement, handling and fulfilment costs.
• Working capitalCash committed and inventory held, plus spoilage and unsold stock where we can measure them.
• OperationsFulfilment reliability, sourcing time and settlement workload.
• RetentionRepeat ordering over a defined period.

For the store owner, the question is not simply whether more orders arrive. It is whether those orders can be fulfilled reliably, at an attractive margin, without tying up more cash or adding more operational complexity than the business can support. That is what we need to establish through the pilot.

Success would mean stronger store-level economics, less unnecessary cash tied up in inventory, reliable fulfilment and repeat household orders — without adding operational costs that outweigh the benefits. It could fail if sourcing after the order proves too slow, if settlements cost too much to manage, or if the orders are too small to matter. Once we have a baseline and a defined measurement period, we want to share what we learn, including what does not work.

Digitisation should strengthen the kirana, not replace it

India’s neighbourhood kiranas do not need to become miniature versions of large retail chains. They need practical ways to turn the trust they have built with customers into better-informed purchasing decisions.

At ShopConnect, we are testing whether capturing household demand before procurement can help neighbourhood stores serve customers more reliably, manage inventory more carefully and build stronger businesses. The idea is simple; whether the economics work in practice is what the pilot must establish.

The next chapter of kirana digitisation should be judged not by how many stores adopt another tool, but by whether those tools make the business measurably better for the people who run it.

Work in distribution or FMCG? We would value your perspective too: write to info@nvipani.com.