Quick Commerce Software in India: How to Build 10-30 Minute Delivery With Dark Stores

What is quick commerce, and why does India suddenly run on it?

Quick commerce, or q-commerce, is the business of delivering everyday goods, mostly groceries, snacks, medicines and household items, to a customer's doorstep in 10 to 30 minutes. It sits between a full grocery e-commerce order that arrives the next day and a walk-in kirana purchase that happens instantly. Names like Blinkit, Zepto and Swiggy Instamart have made this speed feel normal in metros and increasingly in Tier 2 and Tier 3 cities.

The engine behind that speed is not magic and it is not a giant warehouse. It is a small, densely packed neighbourhood facility called a dark store, plus a tight layer of software that decides what to stock, how to pick it, and which rider carries it out the door. The right quick commerce software is what separates a business that delivers in 15 minutes at a profit from one that delivers late and bleeds money.

If you run a grocery chain, a supermarket, a pharmacy, a D2C brand or even a growing kirana, you do not need to become the next Zepto to benefit from this model. A single dark store serving a 2 to 3 kilometre radius, run properly, can turn your existing catalogue into a same-hour or same-30-minute service. This guide walks through exactly how that operation is built, piece by piece, in an Indian context, using rupees, GST, WhatsApp and the realities of Indian streets.

How do you set up and zone a dark store?

A dark store is a retail space with no customers walking in. It exists purely to fulfil online orders fast. Because there are no shoppers browsing aisles, you optimise the layout for pickers, not for footfall. A typical Indian dark store runs between 1,500 and 4,000 square feet and stocks 1,500 to 5,000 SKUs, far fewer than a supermarket but chosen carefully around what a neighbourhood actually buys.

Location is the first decision, and it is the one most people get wrong. A dark store must sit inside the delivery zone it serves, ideally within a 2 to 3 kilometre travel radius so a rider can reach the farthest customer and return inside the promised window. Rent on a side lane costs far less than a main-road retail spot, and since no customers visit, visibility does not matter. What matters is road access for riders and proximity to dense residential clusters.

Inside the store, you zone shelves by pick frequency, not by product category. The fastest-moving items, milk, bread, eggs, cold drinks, chips, sit closest to the packing counter so pickers walk the least. Slow movers go to the back. Cold storage and a chiller handle dairy, frozen goods and, if you are a pharmacy, temperature-sensitive medicines. Every shelf and bin gets a location code so the software can direct a picker straight to the right spot.

• Zone A (hot zone): top 20% of SKUs by order frequency, kept nearest the packing station.
• Zone B (medium): regular sellers, one aisle back.
• Zone C (cold): dairy, frozen, chocolates and any temperature-sensitive stock.
• Zone D (slow and bulky): rice bags, oil tins, cleaning supplies at the rear.
• Staging area: a counter near the exit where packed orders wait for riders.

Once zoning is right, the pick-pack workflow rides on top of it. The moment an order lands a clock starts, and the target is to have it picked, packed and handed to a rider in under 3 to 4 minutes so the rest of the window is left for the ride. The system checks stock, reserves those units so no other order can claim them, and generates a pick list sorted by shelf location rather than the order the customer typed items in. The picker walks one optimised path, scanning each barcode into the tote, and scanning is the quality gate that stops a wrong SKU or an expired batch from ever reaching the bag. The packer then bags it, adds cold-chain packaging, prints the label and stages it under the rider's name. At peak, the software batches pick runs across nearby orders or splits a large order across two pickers, which is the difference between a store that copes with a Sunday-evening rush and one that collapses under it.

How are riders assigned, batched and tracked?

Rider assignment is where quick commerce economics are won or lost. Assign badly and you either keep customers waiting or send half-empty riders on single drops that never pay for themselves. Good quick commerce software assigns each order to the rider who can deliver it fastest while keeping every rider productive.

The system looks at which riders are free, where they are, how far each pending order is, and whether two or three orders are heading in the same direction. When they are, it batches them into one trip so a single rider drops three parcels on one loop instead of making three separate runs. It weighs the delivery promise against efficiency, because batching too aggressively makes the last customer in the batch wait too long. Getting that balance right is a core job of the software, not the dispatch manager guessing.

Once a rider picks up, live tracking kicks in. The customer sees the rider moving on a map with a realistic ETA, your operations team sees every rider on one dashboard, and the system flags any delivery that is running late before the customer complains. GPS also creates a record for proof of delivery and for settling any dispute about whether an order actually arrived.

• Auto-assignment: nearest available rider, weighted by current load and delivery promise.
• Batching: combine 2 to 3 same-direction orders into one trip during peak demand.
• Live tracking: real-time rider location and ETA shared with the customer.
• Proof of delivery: OTP or photo confirmation logged against each order.
• Rider payouts: per-order and per-kilometre earnings calculated automatically for settlement.

How do pin-code serviceability and live inventory work together?

Two invisible systems decide whether a customer can even place an order: serviceability and real-time inventory. Get either wrong and you take orders you cannot fulfil, which is the fastest way to lose a customer permanently.

Serviceability is the rule that says which pin codes, or more precisely which delivery polygons drawn around your dark store, you actually serve. A customer outside the radius should see a polite 'we don't deliver here yet' message rather than a false promise. Well-built software lets you draw these zones on a map, set different delivery windows or minimum-order values per zone, and switch a zone off instantly if a store is overwhelmed or short-staffed. This is the control that keeps you from over-promising during a festival rush.

Real-time inventory is the other half. In quick commerce there is no buffer, if the app says two packets of paneer are in stock, exactly two must physically be on the shelf. Every pick decrements stock the instant it is scanned, so the storefront never sells the third packet that does not exist. For perishables and medicines you also track batch numbers and expiry dates, so the system picks the nearest-to-expiry stock first and blocks anything past its date from being sold. For a pharmacy this is not optional, it is a compliance requirement.

The same data that keeps inventory honest also lets you forecast demand and cut wastage, which matters because dark stores hold a small, perishable-heavy catalogue where guessing wrong is expensive both ways. Overstock milk and vegetables and you bin money every evening; understock the Sunday-evening favourites and you lose the order and the customer. Good software studies your own sales history to find the patterns a manager feels but cannot quantify: weekend SKU spikes, how rain shifts the mix, how a salary week around the first of the month lifts basket sizes, and how festivals like Diwali or Bihu reshape demand for a fortnight. It then suggests how much to stock and when to reorder, and layered with expiry tracking it nudges you to discount or bundle items nearing expiry before they become dead stock. For an Indian grocer on thin margins, cutting perishable wastage from, say, 8% to 3% can be the whole difference between profit and loss.

How do customers order: app, website or WhatsApp?

In India, the ordering surface is not one channel, it is several, and your quick commerce software should feed them all from one catalogue and one stock pool. Force customers into a single app and you shut out the large share of Indian shoppers who prefer to just message a shop.

The core storefront is a customer app and a mobile website where people browse categories, search, add to cart, apply a coupon and pay by UPI, card or cash on delivery. That covers the customers who like a full catalogue experience. But WhatsApp ordering is what unlocks the rest of the market, especially loyal neighbourhood customers and older shoppers. A customer sends a message or a list, an automated assistant confirms items, price and slot, shares a payment link, and sends order and delivery updates, all inside the chat they already use every day. No app to download, no account to create.

Behind every channel sits the same live inventory and the same serviceability rules, so a WhatsApp order and an app order compete for the same paneer packet correctly. Retention tools then bring customers back: loyalty points, reorder-your-usual prompts, and personalised offers pushed over WhatsApp or the app. In quick commerce, where a customer might order three or four times a week, retention is worth far more than any single sale, so this layer directly drives your economics.

• Customer app and mobile-first website with UPI, cards and cash on delivery.
• WhatsApp ordering with an automated assistant for confirmations, payment links and updates.
• One shared catalogue and one live stock pool across every channel.
• Loyalty points, reorder prompts and personalised offers to lift repeat orders.
• Order and delivery notifications on the channel the customer already uses.

Should you integrate with ONDC and aggregators like Blinkit, Zepto or Instamart?

You do not have to choose between running your own storefront and appearing on the big platforms. The smartest Indian operators do both, and the software layer is what makes that manageable instead of chaotic.

ONDC, the government-backed Open Network for Digital Commerce, lets a small seller list once and become discoverable to buyers across many apps on the network, without paying the steep take-rates or being locked into a single marketplace. For a local grocer or D2C brand, ONDC is a genuine channel to reach demand you could never generate alone, and quick commerce software that speaks ONDC's protocol lets you plug in without building the integration yourself.

Aggregators like Blinkit, Zepto and Swiggy Instamart are the other route. Listing your dark store or brand on them buys instant reach into their huge user base, at the cost of commission and less control over the customer relationship. The operational risk is running these as separate silos, with stock and prices drifting apart until you oversell on one channel while sitting on inventory in another. The fix is a single system where your own app, WhatsApp, ONDC and each aggregator all draw from one inventory and one price book, and every order, wherever it came from, flows into the same pick-pack-deliver workflow. That unified control is exactly what an ERP built for commerce provides.

Do the unit economics of quick commerce actually work?

This is the question that matters most, because quick commerce has a reputation for burning cash. It burns cash when operators chase growth by subsidising every order. It works when you understand the economics per order and design your operation to be profitable at a realistic average order value.

Every order carries roughly the same set of costs: the delivery cost of paying the rider, the picking and packing labour, the packaging, the payment-gateway fee, a share of dark-store rent and staff, and perishable wastage. Against that you earn the gross margin on the basket plus any delivery fee you charge. The single biggest lever is average order value, because most of your per-order cost is fixed regardless of basket size. A ₹200 order and a ₹600 order cost almost the same to deliver, but the ₹600 order carries three times the margin to cover it.

That is why serious operators push AOV with minimum-order thresholds, smart bundling and combo offers, and why rider batching matters so much, splitting one rider's cost across three drops instead of one transforms the maths. Tight zoning that cuts picking time, sharp forecasting that kills wastage, and retention that lowers the cost of getting each repeat order all compound into the difference between profit and loss. None of these are heroics, they are settings and disciplines your software should make easy to run and easy to measure.

• Per-order costs: rider payout, pick-pack labour, packaging, gateway fee, rent share, wastage.
• Per-order revenue: basket gross margin plus any delivery fee charged.
• Biggest lever: average order value, since most per-order cost is fixed.
• Push AOV with minimum-order values, bundles and combos.
• Cut cost per order with rider batching, tight zoning and low wastage.
• Track contribution margin per order, not just revenue, from day one.

Ready to build your quick commerce operation the right way?

Quick commerce rewards operators who treat it as a systems problem, not a speed race. Dark-store zoning, disciplined pick-pack, smart rider batching, live inventory, honest serviceability, real forecasting, multi-channel ordering and clear unit economics have to work as one connected machine. Bolt them together from separate tools and the gaps quietly eat your margin. Run them from one platform and speed becomes profitable.

That is what TheManki builds. Manki Commerce ERP is the backbone: it runs your dark stores, real-time and batch-tracked inventory, pin-code serviceability, pick-pack workflows, rider assignment and batching, demand forecasting and one unified view across your own channels, ONDC and aggregators. Manki Commerce AI powers the customer-facing storefront and WhatsApp ordering, so your app, website and chat all sell from the same live stock with personalised, retention-driving experiences. Everything is designed for Indian realities, GST, UPI, Tally-friendly accounting, WhatsApp-first customers and the margins Indian grocers and D2C brands actually work with.

If you are a kirana, grocery chain, supermarket, pharmacy or D2C brand thinking about 10 to 30 minute delivery, start with a conversation rather than a leap. Book a free strategy call with TheManki on WhatsApp at +91 70022 08642 and we will map your catalogue, your delivery radius and your numbers into a quick commerce operation that is built to make money, not just to move fast. Engineering business evolution is what we do.