A dark store is a retail-format building closed to the public and used entirely for fulfilling online orders. It looks like a supermarket inside — aisles, shelves, chillers, familiar product adjacencies — and no customer ever walks through it. Staff pick orders from those aisles for home delivery or collection.
The format exists because of a specific problem. Grocery delivery in under an hour cannot be served from a distribution centre forty miles away, and it cannot be served efficiently from a trading shop full of customers. A dark store puts supermarket-style stock close to where people live, laid out for pickers rather than shoppers.
Why it is called a dark store
The name refers to being closed to customers, not to the lighting. A dark store is “dark” in the sense that its doors are shut to the public and its windows are usually covered — it is a shop with the shutters down, working full-time on orders placed elsewhere.
The lights inside are very much on. Pickers need to read labels and check dates, and many dark stores run longer hours than the shops they resemble.
The term borrows from an older usage in retail: a store that has gone dark has stopped trading. The difference is that a dark store has stopped trading in person whilst trading more than ever online.
Why dark stores have grown
Quick commerce changed the promise. Once one operator offered grocery delivery in fifteen or thirty minutes, the expectation spread. That window is a logistics constraint before it is a commercial one: it forces stock to sit within a few miles of the customer.
Online grocery volumes stepped up and stayed up. The pandemic accelerated adoption sharply, and whilst the peak receded, the baseline settled well above where it was. Retailers found themselves fulfilling a volume of online orders their shops were never designed for.
Picking in a trading shop does not scale. In-store picking works at low volume and degrades quickly. Pickers compete with customers for aisle space, shelves get emptied by online orders in front of shoppers who wanted the same item, and pick rates are far lower than in a purpose-arranged building.
Retail space became available. Store closures and falling footfall left retailers holding leases on buildings in exactly the residential locations that fast delivery requires. Converting one to a dark store uses an asset that was otherwise a cost.
Distribution centres are too far away. A large automated grocery DC is efficient per order and structurally unable to deliver in thirty minutes to most of its catchment.
Dark store vs retail store vs warehouse
| Dark store | Retail store | Warehouse / DC | |
|---|---|---|---|
| Open to customers | No | Yes | No |
| Typical size | Small to medium, shop-scale | Shop-scale | Large to very large |
| Location | Urban and residential, near customers | Shopping streets and retail parks, near footfall | Edge of town, near motorways, land is cheaper |
| Layout logic | Pick speed — fastest lines nearest dispatch | Shopper experience — promotions, browsing, impulse | Storage density and bulk throughput |
| Range held | Narrower; the lines that actually sell online | Broad, including slow sellers for choice | Very broad, often full range |
| Fixtures | Shop shelving and chillers, arranged for picking | Shop shelving arranged for shopping | Racking, bulk storage, mechanised handling |
| Delivery radius | A few miles, often under 30 minutes | Not applicable | Regional or national |
| Staffing | Pickers and drivers | Sales, checkout, replenishment | Warehouse operatives, forklift drivers |
| Order profile | Many small multi-line orders | Individual shopping baskets | Pallets and cases to stores or large orders |
| Cost per order | Higher — small scale, urban rent, fast delivery | Not comparable | Lowest — scale and automation |
The useful way to read that: a dark store is a shop’s format and location running a warehouse’s job. It borrows the proximity and the familiar layout from retail, and the picking discipline from a fulfilment centre — and it inherits the cost problems of both.
How a dark store works
1. The order arrives through an app or website, with a delivery or collection slot.
2. The system routes it to the dark store covering that delivery area, checking stock is available locally.
3. Picking. A picker works from a handheld device along a route optimised for the building, usually picking several orders at once into a trolley of totes — cluster picking by another name. Chilled and frozen lines are picked last to limit time out of temperature.
4. Substitutions. Grocery differs from most fulfilment here: where an item is unavailable, the picker offers a substitute in the moment, which the customer can accept or refuse on delivery. This is a judgement call thousands of times a day and it drives customer satisfaction more than delivery speed does.
5. Consolidation and staging. Orders are assembled by slot, with ambient, chilled and frozen kept apart until the last moment.
6. Fulfilment, by one of three routes.
- Rapid home delivery — a driver takes several orders on a short local route. The dominant model for quick commerce.
- Click and collect — the customer comes to the building at a booked time.
- Kerbside collection — the customer arrives and the order is brought out to their vehicle. Common where the site has parking.
7. Replenishment. Stock arrives from the retailer’s normal distribution network, usually daily, and is put away to the pick face overnight.
The whole cycle for quick commerce can run under thirty minutes from order to doorstep, which is only possible because the stock was already within a few miles and the building was arranged for picking.
Who operates dark stores
Supermarket groups run them at scale. Tesco, Sainsbury’s, Asda and their equivalents across Europe and North America operate dedicated online fulfilment sites alongside in-store picking, and several have converted underperforming shops into dark stores serving dense urban catchments.
Amazon has used the format through Amazon Fresh, operating fulfilment sites for grocery delivery separately from its main network.
Walmart and Kroger in the United States have both invested heavily — Kroger through its partnership with Ocado, building automated customer fulfilment centres, and Walmart through store conversions and dedicated fulfilment space attached to existing shops.
Ocado is a distinct case: it built an online-only grocery model from the start, with no shops at all, and now licenses its automated fulfilment technology to retailers worldwide.
Instacart and similar platforms operate a hybrid, picking from partner retailers’ shops rather than from dedicated buildings, and increasingly from dedicated space within them.
Quick-commerce specialists — Getir, Gorillas, Zapp, Flink and others — built businesses entirely on the format, with small urban sites promising delivery in minutes. That sector has consolidated heavily, and several operators have withdrawn from markets or merged, which is itself instructive about the economics.
Non-grocery retailers use the format too. Pharmacy, pet supplies, DIY and general merchandise all run dark or partially dark sites where local delivery matters.
The benefits
Speed. Stock a few miles from the customer is what makes sub-hour delivery possible at all. No amount of efficiency in a distant DC substitutes for proximity.
Much better pick rates than in-store. No customers in the aisle, a layout arranged for picking, and staff doing nothing else. Pickers in a dark store are considerably more productive than the same staff picking in a trading shop.
Higher availability for online customers. Stock is not being depleted by shoppers at the same time, and the range held is the range that actually sells online.
It protects the shopping experience. In-store picking irritates customers — pickers block aisles and take the last of an item in front of someone reaching for it. Moving that work out resolves it.
It repurposes an asset. A shop with falling footfall in a residential area is close to ideal as a dark store: right location, right size, right fixtures, and the lease is already signed.
Longer operating hours. No trading hours restrictions in the same way, so picking and replenishment can run overnight.
Local inventory visibility. Knowing what is available in each local site enables accurate promises at the point of order rather than apologies afterwards.
The challenges
Profitability is the central problem, and it is not solved. Small-basket grocery delivered in minutes carries a cost per order — picking labour, urban rent, a driver for a handful of orders — that frequently exceeds the margin on the basket. The quick-commerce sector’s consolidation and withdrawals are the visible evidence. Many operators have responded by raising minimum baskets, adding delivery fees, or lengthening the promised window.
Property cost. Dark stores need to be where people live, which is where rent is highest and space is smallest. That is the opposite of the trade-off a conventional warehouse makes.
Low density. Shop-format shelving stores far less per square metre than racking. You are paying urban rent for warehouse-inefficient storage.
Staffing. Pickers and drivers in urban locations, at volumes that swing sharply by hour and by day. Scheduling to demand that peaks at teatime and at weekends is genuinely difficult.
Range limitations. A small local site cannot hold everything, so either the range is cut or some orders have to be split or sourced elsewhere — which destroys the economics of the individual order.
Replenishment complexity. Frequent small deliveries to many small sites costs considerably more per unit than supplying a few large ones.
Perishability. Grocery dark stores hold short-dated stock in a building sized for fast turnover. Getting the forecast wrong produces waste quickly.
Local objections. Sites in residential areas generate delivery traffic, riders and noise, and several cities have introduced planning restrictions on converting retail units to fulfilment use.
The economics, roughly. A useful way to see the problem: a picker assembling around 100 items an hour, an average basket of 25 items, and a driver managing four or five drops an hour in an urban area. That puts perhaps 15 minutes of picking labour and 13 minutes of driver time into a single order, before rent, vehicle cost, packaging or waste. Against a grocery gross margin measured in a few dollars per basket, the gap is obvious — and it is why almost every lever operators have pulled since (minimum baskets, delivery fees, longer windows allowing more drops per run, automation) attacks one of those two numbers.
Where dark stores are heading
Micro-fulfilment is the convergence point. Rather than shop shelving in a converted unit, operators are installing compact automated systems — robotic storage and retrieval in a footprint of a few hundred square metres — either standing alone or bolted onto the back of a trading shop. This directly attacks the two weaknesses of the format: low storage density and high picking labour.
Attached rather than standalone. Building a small automated fulfilment area onto an existing supermarket shares the rent, the stock and the replenishment with a shop that is trading anyway, which is a materially better cost base than a dedicated site.
Hybrid formats. Shops that trade normally at the front and fulfil from a closed area at the back. The flexibility to shift the balance as online share changes is worth a lot.
Slower promises, better economics. Several operators have moved from fifteen-minute promises to one- or two-hour windows, which allows batching several orders per driver and transforms the delivery cost per order. Customers have largely accepted it.
Consolidation of the specialists. The standalone quick-commerce model has proved difficult to fund independently, and much of the capacity has been absorbed by established grocers who can share costs across an existing network.
The format itself is not going away — proximity remains the only way to deliver fast. What is changing is the assumption that a dedicated, manually picked building is the right way to achieve it.
Where to go deeper
- What is a fulfilment centre — the format dark stores borrow their job from.
- What is a distribution centre — the facility they sit downstream of.
- Omnichannel fulfilment — where dark stores fit in a multi-route strategy.
- What is order picking — the activity that dominates their cost.
- What is an AS/RS — the automation behind micro-fulfilment.
- What is a warehouse — how the building types compare.
The one-line takeaway
A dark store is a retail-format building closed to the public and used solely to fulfil online orders, placed close to customers so fast delivery is physically possible. It solves proximity and picking productivity, and it inherits high urban rent and low storage density — which is why the format is moving towards automated micro-fulfilment and towards space attached to shops that still trade.
Frequently asked questions
What is a dark store? A retail-format facility closed to the public and used entirely for fulfilling online orders. It is laid out like a supermarket but arranged for picking speed rather than shopping, and it sits close to customers so delivery within a short window is possible.
Why is it called a dark store? Because it is closed to customers — the shutters are down and the windows are usually covered — not because of the lighting inside, which is fully on for pickers reading labels and dates. The phrase borrows from the retail usage of a store that has “gone dark”, meaning it has stopped trading in person.
What is the difference between a dark store and a warehouse? A dark store is shop-sized, sits in a residential or urban area, uses retail shelving, and holds a narrower range for local fast delivery. A warehouse or distribution centre is far larger, sits on cheaper land near transport links, uses racking for density, and serves a region rather than a few miles.
Who uses dark stores? Major supermarket groups including Tesco, Sainsbury’s, Asda, Walmart and Kroger; Amazon through Amazon Fresh; online-only grocers such as Ocado; platforms like Instacart in hybrid form; and quick-commerce specialists such as Getir and Flink — a sector that has consolidated considerably.
Are dark stores profitable? Frequently not, in the standalone quick-commerce form. Small baskets delivered in minutes carry picking, urban rent and driver costs that often exceed the margin. Operators have responded with minimum basket values, delivery fees, longer delivery windows allowing batched drops, automation, and attaching fulfilment space to shops that already trade.
What is the difference between a dark store and a micro-fulfilment centre? A dark store is generally manually picked from retail-style shelving. A micro-fulfilment centre uses compact automation — robotic storage and retrieval — in a small footprint, which stores far more per square metre and cuts picking labour sharply. Micro-fulfilment is increasingly how the dark store concept is being built.
Can customers collect from a dark store? Many offer click and collect or kerbside collection at a booked time, even though the building is closed for browsing. The customer arrives, the order is handed over, and they do not enter the shop floor.
Warehouse Wise is a free, independent library of guides on inventory management, warehousing and fulfilment, written for the people actually doing the work — new warehouse and inventory staff, small and growing retail and ecommerce businesses, and anyone who holds stock and wants to understand it properly. Every guide explains the term, shows a worked example with real numbers where it helps, and stops. We are independent: we do not sell software or equipment, we take no commission on referrals, and our guides are not ranked by who pays us.