Quick commerce glossary

Online shopping that delivers groceries and daily needs in minutes, usually 10 to 30, from small nearby warehouses called dark stores. Blinkit, Zepto and Swiggy Instamart are the best-known apps in India.

Quick commerce is built for small, urgent, frequent buys: milk at breakfast, a charger before a trip, snacks when guests arrive. Instead of shipping from a big regional warehouse, apps keep a few thousand products in dark stores within a couple of kilometres of homes.

In India the main apps are Blinkit (owned by Eternal), Zepto, Swiggy Instamart, Flipkart Minutes, BigBasket’s BB Now, Amazon Now and JioMart. Specialist apps cover fashion, food and pet supplies. The market has grown fast, and estimates of its size differ a lot.

The model needs dense cities, cheap delivery labour and big enough baskets. Profits have been slow. Blinkit turned adjusted-EBITDA positive only in late 2025. Supporters point to jobs and convenience. Critics point to rider safety, food-safety lapses and pressure on local kiranas.

Six apps run 8,241 dark stores across 498 cities, including Amazon Now’s 853. JioMart also delivers from 2,613 Reliance Stores.

Short for quick commerce: groceries and daily needs delivered in minutes from nearby dark stores.

A small warehouse, closed to walk-in shoppers, that stocks daily needs and only packs online orders for homes nearby. Blinkit, Zepto and Instamart deliver from dark stores.

Inside, a dark store looks like a cramped supermarket. There are aisles of groceries, chillers for milk and ice cream, and the fastest-selling items sit closest to the packing table. Only pickers and riders go in. It is “dark” because the public cannot walk in.

Apps such as Blinkit, Zepto and Swiggy Instamart deliver from these stores, usually in 10 to 30 minutes. Each store covers only a few kilometres, so an app can be live in your city and still not deliver to your society. Newer stores are bigger, often above 5,000 sq ft against 2,000–3,000 sq ft earlier, so they can stock electronics and beauty products too.

A dark store keeps food, so it needs an FSSAI licence, and state food-safety officers can inspect it. Several stores had their licences suspended in 2025 and 2026.

Six apps run 8,241 dark stores across 498 cities. Bengaluru has the most, with 926. Blinkit alone runs 190 there.

A small local warehouse that keeps stock close to customers so orders can go out in minutes. Flipkart and Amazon use the term for what Blinkit and Zepto call a dark store.

The idea is the same as a dark store: keep a few thousand products near dense housing so the trip to your door is short. Only the label changes. Flipkart calls its Flipkart Minutes sites micro-fulfilment centres, and Amazon uses micro-fulfilment and “urban fulfilment” centres for Amazon Now. Blinkit, Zepto and Instamart say dark stores.

Some MFCs run inside or behind existing shops. When Amazon Now launched in Belagavi in September 2026, it used two More supermarket outlets as delivery points, each covering about 15 km. A dark store in a big city usually covers only 2–3 km.

Because the words overlap, store counts from different companies are not always like for like. A partner supermarket and a purpose-built dark store can both count as one “centre”. Amazon Now uses some partner More stores as dark stores, so they count as Amazon Now dark stores. JioMart delivers from Reliance Stores, which stay out of any dark-store total.

Flipkart Minutes crossed 1,000 micro-fulfilment centres in June 2026, across 130+ cities. It now has 1,202 stores. Amazon Now has 853 dark stores, including partner More stores.

A large warehouse on the edge of a city. Brands deliver bulk stock there, and it sends smaller daily loads to many dark stores.

A dark store is too small to take a full truck from every brand. So apps run bigger warehouses, called mother hubs, usually in industrial belts outside the city. Brands deliver to the hub. The hub splits the stock into small lots and sends vans to each dark store, sometimes more than once a day for milk, bread and vegetables. Zepto, for example, runs 75 larger warehouses behind its dark stores.

A mother hub and a dark store do different jobs. The hub takes full trucks, stores bulk stock and ships to stores. The dark store keeps a smaller stock close to homes, packs single orders and hands them to riders. You never order from a hub directly, and the app does not show it to you.

The hub shapes what you see in the app. If an item runs out at the hub, every dark store it feeds runs short. Food-safety rules apply to hubs too. In August 2026, Karnataka’s food-safety department sealed a Zepto warehouse in Hoskote, on the edge of Bengaluru, run with a logistics partner.

Bigger dark stores need less hub space behind them. A 2,500–3,000 sq ft store needs about the same area again in warehouse space. A 7,000 sq ft store needs only about 0.6 times its own size, because it keeps more of its own spare stock.

In July 2026, Blinkit raised its expected cost per store, including its share of warehouse space, to ₹2.5 crore from ₹1 crore. Most of that warehouse share is mother-hub space.

Whether an app can deliver to your address right now. “Not serviceable” means no dark store is close enough, or the nearest one is closed or too busy.

Apps check serviceability address by address. Your phone location or saved address is matched to the nearest dark store. If you are outside that store’s zone, the app shows “not serviceable” or offers slower delivery.

It changes during the day. At peak hours, in heavy rain or late at night, an app may shrink a store’s zone, pause orders or show longer delivery times. That is why the same address can work at noon and fail at midnight.

It also changes over time. When an app opens a new store nearby, an area that was not served can switch on overnight. Store locations on the city and area pages show where an app is likely to deliver.

Companies often describe their reach in PIN codes. Flipkart says Flipkart Minutes covered 8,000+ PIN codes in June 2026.

For you, the PIN code is only a starting point. A large city PIN code can stretch several kilometres. If the only dark store sits at one end, homes at the other end may still be out of reach. The reverse also happens: a store just across the boundary may deliver to you.

Nearly 9 in 10 new dark stores open in PIN codes that already have one. Apps are filling gaps inside cities they already serve more than reaching new areas.

Indore, a tier-2 city, has 50 dark stores. With so few stores in a large city, many outer colonies fall outside every app’s zone.

The list of six-digit PIN codes where an app delivers. It is only a rough guide, because one PIN code can be partly inside and partly outside a store’s delivery zone.

How far from a dark store an app will deliver. In big cities it is usually about 2–3 km, which is what makes 10 to 20 minute delivery possible.

A short radius is what makes fast delivery work. In early 2025, Zepto’s Tamil Nadu stores each served a 2–3 km radius, so riders could keep to about 15–20 km/h. In January 2026, Eternal’s Deepinder Goyal said the average Blinkit trip was about 2 km.

The zone is not a neat circle. Apps draw it around roads, one-way streets, rivers, railway lines and gated areas, and shrink it at busy times. So a store 2 km away by air can still be out of range by road.

In tier-2 cities and smaller towns, where stores are few, the radius can be much larger. When Amazon Now launched in Belagavi in September 2026, it used two supermarket outlets, each covering roughly 15 km. Deliveries there take longer than in a metro.

Gurugram has 282 dark stores, so most sectors have several stores within 2–3 km. Belagavi has 10.

The final leg of delivery, from the store to your door. It is usually the costliest part of getting an order to you.

Goods travel cheaply in bulk: by truck from the factory to a warehouse, then by van to a dark store. The last mile is different. Each trip carries one small order to one address. Rider pay, fuel and waiting time make it a big share of the cost of every quick delivery order.

Dark stores exist to shorten the last mile. With stock kept within 2–3 km of homes, each trip takes only a few minutes, so one rider can make more deliveries in an hour.

The last-mile cost is why apps care so much about basket size. A ₹600 basket and a ₹60 basket cost about the same to deliver. Small-order fees and free-delivery limits are really about paying for the last mile.

In January 2026, Eternal said the average Blinkit order travelled about 2.03 km, with roughly eight minutes of riding. Short trips like that are what make the dark-store model work.

The promise, made popular by Blinkit and Zepto, that your order arrives about 10 minutes after you place it. In January 2026 the government asked apps to stop advertising it.

“10 minutes” was a marketing line, not a guarantee. Real delivery time depends on how close the nearest dark store is, how busy it is, traffic and weather. In big cities most orders still arrive in 10–20 minutes.

Gig worker unions said the promise pushed riders to drive dangerously. After rider strikes on 25 and 31 December 2025, Labour Minister Mansukh Mandaviya met the apps on 13 January 2026 and asked them to drop 10-minute claims. Blinkit changed its tagline to “30,000+ products delivered at your doorstep”.

The minister made a request, not a law. No rule bans fast delivery, and apps still show delivery times in minutes.

In January 2026, Eternal said the average Blinkit trip was about 2 km, with roughly eight minutes of riding. Its argument: short distances, not rash riding, make the times possible.

Delivery in a time window you choose, such as tomorrow 7–9 am, sent from a large warehouse instead of a nearby dark store. It is the older way of buying groceries online.

Before quick commerce, online grocery in India was mostly slotted. BigBasket and Amazon Fresh collected orders, picked them in big warehouses and delivered in batches. Batching keeps costs low, because one van serves many homes in one trip.

Slotted delivery suits a weekly or monthly shop with a big basket. Quick delivery suits urgent top-ups. Many homes use both, and several companies offer both. BigBasket runs slotted delivery alongside BB Now, and Amazon offers same-day and next-day delivery alongside Amazon Now.

The line between the two is blurring. As dark stores get bigger and carry more products, larger baskets are moving to quick delivery apps. Slotted services, in turn, are cutting their delivery windows.

On BigBasket’s scheduled orders in most cities, non-members paid no delivery fee above ₹1,000, ₹10 between ₹600 and ₹1,000, and ₹50 below ₹600 (October 2026). So a ₹1,200 monthly shop went free, while a ₹400 top-up cost ₹50 extra. Amazon, for its part, has over a million products for same-day delivery and over four million for next-day delivery, alongside tens of thousands on Amazon Now.

A kitchen that cooks only for delivery orders, with no seating or walk-in counter. It is to restaurants what a dark store is to shops.

Cloud kitchens grew in India with Swiggy and Zomato food delivery. One kitchen can run several brands, each with its own listing on the apps, from a low-rent spot away from the main road.

Quick delivery has borrowed the idea for fast food. Startups such as Swish run small kitchens close to customers and promise food in 10–15 minutes. Blinkit, Zepto and Swiggy have also tried ready-to-eat menus made next to or inside dark stores.

The rules are the same as for any food business. A dark kitchen needs an FSSAI licence and local trade and fire approvals, and it can be inspected. Unlike a dark store, it cooks food, so inspectors check cooking hygiene, not just storage.

Swish raised about $14 million in March 2025 for 10-minute food in Bengaluru, then $38 million in March 2026 to expand to more cities. Its kitchens work like dark stores for cooked food: small, close to homes and closed to walk-in customers.

A mid-sized Indian city, a step below the big metros. Lists vary, but Indore, Jaipur, Lucknow and Coimbatore are usually counted as tier 2.

There is no single official list for shopping and retail. The government’s HRA city classes and several bank and market lists group cities by population and economic weight. On city pages, each city gets a tier from 1 (metros) to 3 (smaller towns), so coverage can be compared.

Quick delivery started in the metros and is now moving down the tiers. About 75% of Blinkit’s new stores in mid-2026 opened outside the metros. Flipkart said Flipkart Minutes grew 42 times in tier-2 and tier-3 markets in a year, and Amazon Now is in more than 120 cities.

In smaller cities, stores are fewer and further apart. Coverage can be patchy beyond the central areas, and delivery takes longer than in a metro.

42 cities with dark stores count as tier 2, with 1,711 dark stores between them. Lucknow has 171 and Indore has 50.

Dark-store staff who collect the items in your order from the shelves (picking) and bag them for the rider (packing). In many stores one person does both.

Besides orders, pickers receive stock from the mother hub, put it on shelves, check expiry dates and remove damaged goods. The expiry check is the one shoppers feel. Expired products on shelves were among the most common findings in food-safety raids on dark stores in 2025 and 2026.

These are usually shift jobs, often hired through staffing agencies. Pay, shift hours and how to apply differ by app and agency, and the dark store jobs guide compares them. Never pay anyone to get one of these jobs.

When you order, it pops up on a handheld device or screen in the dark store. The picker walks a set route through the aisles, scans each item and brings the basket to the packing table. The packer checks the items, bags them, adds the bill and hands the bag to a waiting rider. Apps aim to go from order to handover in about two minutes.

Dark-store staff who check and bag each order and hand it to the rider. Covered with pickers, since one person often does both jobs.

The person who brings your order from the dark store to your door, usually on a motorbike or e-bike. Apps call riders delivery partners because most are not employees.

Riders on quick delivery apps are gig workers. They log in to the app, wait near a dark store and get paid per order plus incentives, not a fixed salary. Most use their own bike and pay for fuel, so take-home pay is lower than the headline figures.

Since the labour codes took effect on 21 November 2025, gig and platform workers are legally recognised, and apps must pay 1–2% of annual turnover into a social security fund. Draft rules from December 2025 set how many working days a rider needs to qualify for benefits.

Rider safety drove one of the biggest changes of 2026. After rider strikes on 25 and 31 December 2025, the Labour Minister asked apps to stop advertising 10-minute delivery, and Blinkit changed its tagline.

Under the December 2025 draft rules, a rider who works 100 days in a year for one app clears the 90-day mark and qualifies for benefits. A rider who works 60 days for one app and 50 for another has 110 days in all, short of the 120 needed across several apps. Keep a record of your working days on each app.

A small charge per order that some quick delivery apps add at checkout for picking and packing your items. It is separate from the delivery fee and the platform fee.

By 2025, handling fees were common. Apps used them to earn more per order without raising product prices. The fee is usually a few rupees, shown as its own line at checkout, with 18% GST on top.

A handling fee is not the same as a platform fee. In March 2024, Zepto added a ₹2 platform fee alongside a separate ₹15 late-night handling fee, so one bill could carry both.

Shoppers complained that fees showed up only late in checkout, which regulators treat as drip pricing. A fee shown upfront is legal. Zepto removed its handling, surge and convenience fees in November 2025.

Take a ₹150 cart on Instamart at November 2025 rates. A cart under ₹199 carried a ₹30 delivery fee, and a cart under ₹299 carried the ₹9.8 handling fee. So the bill came to ₹189.80 before GST on the fees, over 26% more than the items alone. Add ₹150 more to cross ₹299 and both fees went away. These amounts can change without notice.

A flat charge of a few rupees per order that an app adds for using its platform. It is separate from delivery, handling and small-cart fees.

A platform fee does not pay for any one service. The app adds it to every order, or to some users’ orders, to earn a little more per order. Food delivery apps made it familiar, and quick delivery apps copied it.

Zepto was the first quick delivery app to charge one, from March 2024. Blinkit and Instamart charged no such fee on grocery orders then, though Zomato and Swiggy did charge one on food delivery.

Some apps now use “no platform fee” as a selling point.

Like any mandatory charge, a platform fee must be shown before you pay. Under the 2023 dark-pattern guidelines, a mandatory charge that appears only on the last screen counts as drip pricing.

In March 2024, Zepto’s ₹2 platform fee sat alongside a ₹15 late-night handling fee, so a late order could carry two small charges plus delivery. In September 2026, Flipkart’s new Flipkart Minutes app had no platform fee and free delivery above ₹99. Fees change often, so check the bill in the app.

A small extra charge some apps add per order, named for the convenience of ordering. On a quick delivery bill it works much like a platform or handling fee.

Apps use different names for similar small charges. Bistro, Eternal’s quick food app, began charging a ₹5 convenience fee in June 2025, shown only in the final bill summary. Zepto removed its handling, surge and convenience fees in November 2025.

An extra charge on orders below a minimum value, often ₹99. It covers part of the cost of delivering a very small order, or nudges you to add more.

Delivering a ₹60 order costs an app about as much as a ₹600 one: a picker, a bag and a rider trip. A small-cart fee makes you pay part of that cost, or pushes you to add items and cross the limit.

The limit is often ₹99, though it varies. Above it, you may still pay a delivery fee until you reach the free-delivery level, commonly ₹199 or ₹299. Some apps have dropped small-cart fees during fee wars. In November 2025, for example, Zepto moved to a flat ₹30 delivery fee on orders below ₹99, with no separate small-cart charge.

Say you order ₹80 of milk and bread on Blinkit at November 2025 rates. The cart is under ₹99, so a ₹20 small-cart fee applies. It is also under ₹199, so the ₹30 delivery fee applies too, plus the ₹4 handling fee. You pay ₹134 before GST on the fees, about 68% more than the items cost. Adding ₹20 of items removes only the small-cart fee. Crossing ₹199 removes the delivery fee as well.

A temporary extra charge an app adds when demand is high or riders are short, such as at dinner time, late at night or on festival days.

Surge pricing comes from ride-hailing apps. When many people order at once, the app adds a fee to pay riders more and to slow demand a little. On quick delivery apps it shows up as a peak-hour, late-night or high-demand fee on top of the normal delivery fee.

Some apps change the free-delivery limit by time of day instead. Nothing on the bill is called a surge fee, but a peak-hour order needs a bigger basket to ship free, so it works the same way.

A surge fee must be shown before you pay. Under the government’s 2023 dark-pattern guidelines, adding a mandatory charge late in checkout counts as drip pricing. Surges are most likely when a whole city orders at once, such as on festival days.

Zepto dropped surge fees completely in November 2025. In August 2026 it brought back a higher limit at peak hours: free delivery only above ₹299, against ₹199 at other times.

An extra charge some delivery apps add when it rains or the weather is bad. Apps say it pays riders more for working in rough conditions.

Monsoon deliveries are slower and riskier. Roads flood, traffic crawls and fewer riders log in. Some apps add a rain fee to each order and say it pays for extra rider incentives.

Rain fees are taxable. Handling and rain fees on quick delivery orders attract 18% GST, which shows on the bill.

Rain fees were among the delivery charges users named in 2025 complaints about hidden costs on quick delivery apps. BigBasket says its BB Now delivery fee changes with cart value, time of day and weather. A rain fee should appear as its own line before you pay.

Apps do not publish rain-fee amounts, and they differ from app to app, so take a round number. If an app adds a ₹20 rain fee, 18% GST makes it ₹23.60. In Mumbai or Bengaluru, where monsoon showers are frequent between June and September, a fee like that can appear on many days. Ten rainy-day orders a month would add about ₹236. Check the bill line before you place the order.

Showing a low price first, then adding mandatory charges step by step at checkout, so the final bill is higher than the price you first saw.

Drip pricing is one of the dark patterns banned by India’s 2023 guidelines. On a quick delivery app, your basket shows ₹180, then a handling fee, a small-cart fee, a late-night fee and GST appear only on the last screen.

The CCPA’s December 2025 order against Zepto found that mandatory fees such as handling charges showed up late in the purchase flow. Zepto was told to show all charges upfront and prove it had done so within 15 days.

Fees themselves are not illegal. The rules say mandatory charges must be clear before you commit. Optional extras, such as a tip for the rider, must not be pre-selected.

In November 2025, a Blinkit cart under ₹99 carried a ₹20 small-cart fee and a ₹4 handling fee, on top of the delivery fee. Under the drip-pricing rules, charges like these must be visible before you reach the final payment screen.

App or website design that tricks or pushes you into something you did not mean to do, such as paying hidden fees or buying an add-on.

India’s Central Consumer Protection Authority (CCPA) issued the Guidelines for Prevention and Regulation of Dark Patterns in 2023. They name specific tricks, including drip pricing, basket sneaking, false urgency and confirm-shaming. Breaking them can lead to penalties under the Consumer Protection Act, 2019.

Quick delivery apps came under the spotlight in 2025. Users complained about fees added late in checkout and paid memberships switched on by default. In November 2025, 26 e-commerce and quick-commerce platforms filed self-audit declarations with the Consumer Affairs Ministry. In December 2025, the CCPA fined Zepto ₹7 lakh for drip pricing and for pre-selecting Zepto Pass.

The guidelines cover every platform selling to Indian consumers, quick delivery apps included. The CCPA can act on complaints or on its own. If you spot a dark pattern, take screenshots and complain through the app and the National Consumer Helpline.

In November 2025, Zepto’s CEO Aadit Palicha said the app’s earlier fee design “was a mistake” and that it had been rolled back within about two months.

A brand that sells straight to you through its own website or app, or through online platforms, instead of relying mainly on distributors and shops.

Indian D2C brands in snacks, drinks, personal care and home products use quick delivery apps as a fast way to reach buyers. A listing on Blinkit or Zepto can put a new brand in thousands of dark stores at once, without building a distributor network city by city.

It comes at a cost. Brands pay commissions and, more and more, for ads and visibility inside the apps. Supplying dark stores also means keeping small amounts of stock in many places, which adds logistics work.

Watch out for agents who promise listings for a fee. In October 2025, Blinkit’s CEO warned brands about people taking money to get products listed. Contact a platform’s seller team only through its official website. Stock also has to move fast. A dark store has limited shelves, and a product that does not sell quickly can lose its place to a rival.

Keep the paperwork ready before you apply. Apps usually ask for a GST registration, a GS1 barcode on every product and, for food, an FSSAI licence. If your stock will sit in the app’s warehouse in another state, you usually need to add that warehouse to your GST registration as an additional place of business (APOB), or register for GST in that state.

In June 2026, Flipkart said Flipkart Minutes works with 500+ D2C brands. In Coimbatore it also stocks local names such as Aavin and Sakthi Masala. Network size changes what a listing is worth: Flipkart Minutes has 1,202 stores, against 2,621 for Blinkit and 877 for BigBasket BB Now. Each app runs its own seller sign-up, so listing on Blinkit, Zepto or Instamart means applying to each one separately.

A single, distinct product in a store’s range. Each size, flavour or pack is its own SKU, so a 500 ml and a 1 litre bottle of the same oil are two SKUs.

One brand of atta sold in 1 kg, 5 kg and 10 kg packs is three SKUs. Add a multigrain version in the same sizes and it becomes six. Shops count their range in SKUs. A neighbourhood kirana carries a few hundred to a couple of thousand. A dark store carries several thousand, and the biggest apps now advertise over 10,000. A large supermarket carries tens of thousands. More SKUs mean more choice, but also more shelf space, more stock to manage and more risk of items expiring.

Apps have been adding SKUs fast, moving beyond groceries into electronics, beauty, toys and home goods. Since January 2026, Blinkit’s tagline talks about “10,000+ products”. The race for a wider range, more than speed, is the main reason new dark stores are getting bigger.

Fresh SKUs such as milk, curd and vegetables have the shortest shelf life. They cause most of the wastage, and most of the expired-stock findings in food-safety inspections.

Blinkit’s inventory losses were about 1.8% of NOV in April–June 2026, roughly ₹308 crore, mostly on perishables. Managing fresh SKUs is one of the hardest parts of running a dark store.

The full range of products a store or app sells, counted in SKUs and categories. A wider range means more choice, but it needs more space and stock.

Early quick delivery was about a small set of daily needs, delivered very fast. By 2026 the fight had shifted to range. Apps now add phones, small appliances, beauty products, toys, pooja items and even ready-to-eat food from in-store cafes.

Range and speed pull against each other. A small store can sit close to homes but holds less. A big store holds more, but costs more and may be further away. New stores of 5,000–7,000 sq ft are replacing the older 2,000–3,000 sq ft format, mainly to widen the range.

Range also differs by city. Apps stock local brands to compete with neighbourhood shops. In Coimbatore, Flipkart Minutes carries Aavin, Krishna Sweets and Sakthi Masala. Regional snacks, sweets and staples mean a store in Chennai can look quite different from one in Lucknow.

Since January 2026, Blinkit’s tagline has promised “30,000+ products delivered at your doorstep”. A neighbourhood kirana carries a few hundred to a couple of thousand SKUs. So one Blinkit store near you can hold five or more times the range of the shop down the road, packed into a space that is often smaller than a supermarket floor.

An item that is not available at a dark store right now. Each store has its own stock, so a product can be out of stock at your nearest store and in stock a few kilometres away.

An extra address added to a seller’s GST registration. If a brand’s stock sits in a platform’s warehouse, the brand usually has to add that warehouse as an APOB, or register for GST in that state.

A unique product barcode (also called EAN or GTIN) issued in India by GS1 India. Quick-commerce platforms usually ask brands for one on every product, so dark stores can scan it.

Inwarding is when the warehouse checks and shelves the stock a seller delivers. The GRN (goods received note) records how many units were accepted, and sellers are usually paid on that count.

The share of units in a platform’s purchase order that the seller actually delivered. A low fill rate can lead to smaller orders or penalties under the seller agreement.

Stock a platform sends back to the seller, such as damaged, near-expiry or slow-selling items, on the terms set in the seller contract.

The total value of all orders placed on an app in a period, at the prices customers paid. Some companies include taxes and customer fees in it.

GOV is the headline “how much was sold” number. Swiggy gives Instamart’s size as GOV every quarter, and most estimates of market size use it too.

GOV is not revenue. When an app runs a marketplace, its revenue is only the commission, fees and ad income it keeps from those orders. Even when the app owns the stock, revenue is counted after GST and discounts. So a company can show GOV in thousands of crores and revenue that looks much smaller.

Definitions vary. Some companies include delivery fees, and some leave out cancelled orders. Compare GOV across companies with care, and always note the period.

Instamart’s GOV was ₹9,230 crore in July–September 2026, up from ₹7,907 crore in April–June. Its contribution margin slipped to −1.0% of GOV, from −0.2% the quarter before.

The total value of goods sold through a platform in a period, before returns, discounts and the platform’s own costs. It is close to GOV, and the two are often used for the same number.

In Indian quick commerce, GMV and GOV are used almost the same way. Strictly, GMV counts the value of the goods, while GOV counts the value of the orders, which can include delivery and other fees. Market estimates often say “GMV” for the same headline sales figure.

GMV is good for sizing a market and tracking growth. On its own it says nothing about profit, because it ignores how much of that value the company keeps. For that, look at take rate, contribution margin and adjusted EBITDA.

When you see a market figure, check three things. First, the year, since Indian financial years run April to March. Second, whether it is GMV, GOV or net order value. Third, whether it comes from the company itself or is an outside estimate.

India’s quick-commerce market was estimated at about US$13 billion in FY26. That is an outside estimate of sales value, not a figure any company published, so quote it with its year and as an estimate.

The value of orders after removing discounts paid by the app and cancelled orders. Eternal uses NOV as the main size measure for Blinkit.

NOV sits between GOV and revenue. It removes money the customer never actually paid, such as discounts the app funded and orders that were cancelled. So it is closer to what customers really spent than a gross figure.

NOV is the right number for tracking Blinkit quarter by quarter. It does not match Instamart’s GOV exactly. A simple “Blinkit is X times Instamart” using the two headline numbers will overstate or understate the gap a little.

Eternal also gives NOV per store per day, which shows how busy each dark store is. It is one of the clearest signs of whether a store network is maturing or being stretched.

Blinkit’s NOV was ₹17,132 crore in April–June 2026, up 86% from a year earlier. Per store, that came to ₹8.27 lakh a day, against ₹7.34 lakh a year before.

The average amount a customer spends per order. You get it by dividing total order value by the number of orders.

AOV tells you how big a typical basket is. If a store takes 1,000 orders worth ₹5 lakh in a day, its AOV that day is ₹500. In quick commerce the number matters a lot. Picking, packing and delivering an order costs about the same for one packet of chips as for a week of groceries. A bigger basket spreads that cost over more rupees of sales, which puts basket size at the centre of quick-commerce unit economics.

That is why apps set free-delivery limits (₹99, ₹199 or ₹299, depending on the app and the time), charge small-cart fees on tiny orders and show “add ₹40 more” nudges at checkout. Each one is meant to raise AOV.

Companies measure order value a little differently. Eternal gives Blinkit’s net order value per order, while Swiggy gives Instamart’s gross order value. Compare like with like before reading much into small gaps.

Blinkit’s average net order value was ₹518 in April–June 2026, down from ₹525 the quarter before. Flipkart Minutes’ average order, leaving out phones, was estimated at ₹500–530 in September 2026.

The share of each order’s value that the app keeps as income, from commissions, customer fees and ads.

Say you buy ₹500 of groceries and the app earns ₹100 from that order through brand commissions, delivery and handling fees, and ads. Its take rate is 20%. A higher take rate leaves more money to pay for riders, dark stores and discounts.

Ad income from brands is now a big part of the take rate. Brands pay to show up at the top of search or on the home screen. Customer fees such as handling, small-cart and late-night fees add to it too, which is why every fee change makes news.

When an app moves from a marketplace to owning its stock, take rate gets harder to compare. The company now books the full sale as revenue and earns a trading margin instead of a commission. Its revenue on paper jumps even if the money it keeps per order hardly changes.

Swiggy said Instamart broke even at the contribution level in May 2026 mainly through “higher per-order monetisation”. It kept a bigger share of each order, which is a rising take rate.

What an app has left from an order after paying the costs tied to that order, such as delivery, packing, payment charges and store running costs. Head-office costs are not counted.

Contribution margin answers one question: does each order make money on its own? If yes, more orders help. If no, growth adds to losses. It is usually shown as a percentage of order value.

Take a ₹500 order that brings the app ₹100 from commissions, fees and ads, while the rider, packing, payment charges and the store’s share of rent and staff cost ₹110. The contribution is −₹10, or −2% of order value. Raise income by ₹15 per order, through a fee or more ad sales, and the same order turns positive at +1%.

Companies count costs a little differently. Most include the rider, picking and packing, and dark-store rent and staff. Most leave out brand marketing, technology and head-office costs. So two companies’ figures are close, but not exactly comparable.

A positive contribution margin is a milestone, not profit. A business can make money on each order and still lose money overall once marketing and head-office costs are added. That gap is the difference between contribution margin and adjusted EBITDA.

Instamart’s contribution margin was −5.6% of GOV at its worst, in January–March 2025. By April–June 2026 it was −0.2%, close to break-even, before slipping to −1.0% in July–September 2026.

Earnings before interest, tax, depreciation and amortisation, with a few items such as employee stock option costs taken out. Companies use it as a rough measure of operating profit.

EBITDA leaves out loan interest, tax and the wear and tear on assets. “Adjusted” EBITDA also removes costs the company sees as non-cash or one-off. In Indian internet companies, the main one is ESOP (stock option) cost. Under Ind AS 116, rent shows up partly as depreciation, so companies also explain how they treat leases.

In quick commerce, adjusted EBITDA is the figure investors watch to see whether a business can pay for itself. It comes after contribution margin, because it also counts marketing, technology and head-office costs.

Read it with care. It is not cash flow and it is not net profit. A company can show positive adjusted EBITDA while spending heavily on new dark stores. That spending is capital expenditure, and it does not show up in EBITDA.

Blinkit turned adjusted-EBITDA positive at ₹4 crore in October–December 2025. It made ₹37 crore in January–March 2026, ₹102 crore in April–June and ₹184 crore in July–September. In the July–September quarter, Instamart’s adjusted EBITDA was a loss of ₹895 crore.

The money made or lost on one unit of business. In quick commerce, the unit is usually one order or one dark store.

Per order, unit economics adds up what the app earns (commission or trading margin, customer fees, brand ad income) and subtracts what it spends (picking and packing, the rider, packaging, payment charges, discounts and a share of store rent). The result is the contribution per order.

Per store, the question is how many orders a store needs each day to cover rent, staff, power and cold storage. A new store usually loses money until it gets enough orders, which can take months.

The numbers shifted in 2026. Stores cost more to build, about ₹2.5 crore for a large store. But bigger stores carry more products, get bigger baskets and handle more orders. By mid-2026 there were also warnings that metros may have more stores than can run profitably.

In July 2026, Blinkit raised its expected steady-state sales per store to ₹11 lakh a day from ₹7 lakh, while also raising its cost per store to ₹2.5 crore. Busier stores are meant to pay for the bigger investment.

How much a dark store handles in a day, counted as orders per day or sales per day. It is the main sign of whether a store is busy enough to make money.

A dark store has high fixed costs: rent, staff on every shift, cold storage and power. The more orders it handles, the smaller the share of those costs each order carries. Throughput is the link between a store’s costs and its profit.

Companies measure it differently. Eternal gives Blinkit’s net order value per store per day, while analysts estimate orders per store per day. Throughput falls when an app opens many stores at once, because new stores start slowly and can take orders from older ones nearby.

Busier stores need enough pickers, packers and riders. When stores get busier, staffing and stock checks can slip. That is one reason food-safety problems show up in fast-growing networks.

In September 2026, Flipkart Minutes stores were estimated to average 800–1,000 orders a day, with mature stores at 1,200–1,500. For January–March 2026, estimates put Zepto at about 2,140 orders per store per day, Blinkit at 1,425 and Instamart at 1,093. Multiply Blinkit’s 1,425 orders by its average order of ₹525 that quarter and you get roughly ₹7.5 lakh of sales per store per day, in the same range as the ₹8.27 lakh Eternal gave for the next quarter.

The number of people who use an app at least once in a week. These figures usually come from app-tracking estimates, not from the companies, so treat them as rough.

Customers who placed at least one order in a month. Companies use names such as MTU or monthly transacting customers, and each defines it slightly differently.

The DRHP is the draft prospectus a company files with SEBI before an IPO. The UDRHP is the updated draft it files after SEBI’s comments, when it took the confidential route.

A DRHP sets out a company’s business, finances, risks and legal cases, and how much money it wants to raise. Under SEBI’s confidential pre-filing route, the first draft stays private. After SEBI gives its observations, the company files an updated draft (UDRHP-I) for public comment, and then a red herring prospectus with the price band.

For quick commerce, these drafts reveal numbers companies do not otherwise publish: orders, stores, cities, losses and pending regulatory cases.

Zepto is the main example. It filed confidentially in December 2025, got SEBI’s observations in May 2026 and filed its updated draft in June 2026, with an ₹8,010 crore fresh issue. On 31 July 2026 it put the IPO on hold, and it now aims to list between February and May 2027.

Zepto’s updated draft showed 1,139 stores on 31 March 2026, FY26 revenue from operations of ₹22,624 crore and about 17.5 lakh orders a day. Zepto now has 1,392 stores.

The food licence or registration that every business making, storing, moving or selling food in India must hold, under the Food Safety and Standards Act, 2006. Every dark store that stocks food needs one.

The Food Safety and Standards Authority of India (FSSAI) makes the rules, and state food-safety departments do most of the inspections. The licence number should appear in the app and on your bill.

Inspectors can visit a dark store like any food shop. They check storage temperatures, expiry dates, pest control, hygiene and records such as staff medical certificates. They can issue improvement notices, destroy unsafe stock, take samples and suspend the licence until problems are fixed.

Food-safety action against dark stores has grown since mid-2025, with cases in Mumbai, Pune, Bengaluru, Hyderabad and Jaipur.

On 30 September 2026, Telangana’s food-safety team (TG SAFE) inspected 16 dark stores in Hyderabad and suspended four FSSAI licences: two at Instamart stores, one at Zepto and one at Flipkart Minutes.

A company that is majority owned and controlled by Indian residents. Under India’s foreign-investment rules, only such a company may sell its own stock to shoppers online.

India allows 100% foreign investment in online marketplaces. It does not allow foreign-owned companies to run inventory-based online retail, where the platform buys stock and sells it to you. A company more than 50% owned or controlled by foreigners counts as foreign-owned.

Listed quick-commerce companies have many foreign shareholders. To own inventory, they cap foreign ownership below 50% and keep Indian residents in control of the board.

By July–September 2025, about 80% of Blinkit’s sales came from its own inventory. Swiggy plans to move Instamart to the same model within two to four quarters of its August 2026 vote.

In a marketplace, independent sellers own the stock and the app earns a commission. In an inventory model, the app buys the stock itself and sells it to you.

For most of their history, Blinkit, Zepto and Instamart ran as marketplaces. The products in a dark store legally belonged to seller companies, often linked to big distributors. The app earned commissions, fees and ad income. The marketplace model kept them inside the foreign-investment rules.

Owning the stock gives the app more control over prices, range and supplier terms, and it earns a trading margin. It also brings more responsibility. The app becomes liable for faulty products, consumer complaints, tax and the FSSAI licence at each store. It also needs more working capital to hold stock.

Under the inventory model, revenue on paper jumps, because the full sale value is booked instead of just the commission. That is an accounting change, not a sudden jump in business.

Blinkit’s adjusted revenue rose 756% year on year in July–September 2025, mostly because it now books the full value of goods sold. By then about 80% of its sales came from its own stock.

India’s rules on foreign investment in online shopping. Foreign money can fully own a marketplace, but a foreign-owned company may not sell its own stock to shoppers online.

Under the government’s FDI policy, 100% foreign investment is allowed in the marketplace model, where the platform links buyers with independent sellers. It is not allowed in the inventory model, where the platform owns the goods it sells to consumers. The policy also limits how much of a marketplace’s sales can come from one seller group, and bars platforms from controlling prices.

That is why quick delivery apps with large foreign shareholders ran as marketplaces with linked sellers. To own stock, a company must be Indian-owned and controlled, which means keeping foreign ownership below 50%.

The government has also signalled that it wants more Indian ownership in quick commerce. In 2025, Commerce Minister Piyush Goyal raised concerns about foreign shareholding in Zepto.

Eternal capped foreign ownership at 49.5% in April 2025 so that Blinkit could own its stock. Swiggy did the same for Instamart in August 2026.

India’s competition regulator. It looks into cartels, misuse of market power by big companies, and mergers that could hurt competition.

The CCI works under the Competition Act, 2002. A case usually starts with a complaint. The commission first forms a preliminary view. If it sees a likely problem, it orders its Director General to investigate, which can take a year or more. It can fine companies and order them to change how they do business.

Quick commerce reached the CCI in 2025. In February 2025, the All India Consumer Products Distributors Federation complained that Blinkit, Zepto and Instamart used deep discounts and exclusive brand deals to push out distributors and kiranas. By April 2025, the CCI had asked for market-share data and evidence of below-cost selling. In June 2025 it was running a preliminary inquiry.

No formal investigation order has been made public so far. Any such order would appear on the CCI’s website.

The complaint named the companies behind three apps: Blink Commerce (Blinkit), Bundl Technologies (Instamart) and Zepto Marketplace. Together, the three apps handled about 197 crore orders in FY26.

The All India Consumer Products Distributors Federation, a trade body of FMCG distributors. It is the main group fighting quick delivery apps over pricing.

Distributors sit between brands and shops. They buy from companies such as Hindustan Unilever or ITC and supply kiranas. Quick delivery apps often buy straight from brands and cut distributors out, which is why the federation campaigns against them.

Its main moves so far: a letter to the CCI in October 2024; a formal complaint on 20 February 2025 alleging predatory pricing, with price comparisons for 100 products; and a letter to SEBI on 30 December 2025 against stock-market listings by loss-making quick-commerce companies. It has also written to the finance ministry and FSSAI.

Its members, the distributors and stockists who supply neighbourhood shops, say their sales fall when brands sell to apps at lower prices. Figures the federation gives, such as numbers of kirana closures, are an industry body’s claims. They have not been independently checked.

The SEBI letter was aimed mainly at Zepto. Zepto had filed its draft prospectus confidentially in December 2025, and it received SEBI’s observations in May 2026.

A neighbourhood grocery shop, usually family-run, that sells daily needs. Many give regular customers credit and deliver within the colony.

Most Indians still buy groceries at a kirana. It is close by, the owner knows you, gives credit and takes orders on the phone or WhatsApp. Many deliver within the colony, which was quick delivery long before the apps.

Distributors and kirana groups say app discounts hurt these shops. The All India Consumer Products Distributors Federation complained to the CCI in February 2025 about predatory pricing, and the CCI began a preliminary inquiry in June 2025. In Meghalaya, a tribal district council refused Blinkit and Instamart trading licences in 2026, pointing to more than 4,000 local grocery stores in Shillong.

Some apps work with kiranas instead. KiranaPro sent orders to nearby shops, and Amazon Now has used supermarket outlets in smaller cities. Others compete with the kirana head-on, on price and speed. Either way, the local shop is still the benchmark apps measure their convenience against, especially in smaller cities.

Jaipur is a city of over 30 lakh people, with 129 dark stores. Neighbourhood kiranas still far outnumber them.