Channel design & omnichannel

A channel is the route a product takes to reach the person who buys it. An Indian FMCG brand typically sells through several at once:

• General trade: lakhs of small kirana stores, reached through distributors and stockists who visit, take orders and deliver small, frequent drops on credit.
• Modern trade: supermarket chains that order large pallet loads to their own warehouses, with strict delivery windows and penalties.
• E-commerce marketplaces: orders arrive at the marketplace’s fulfilment centres in bulk, then ship as single parcels.
• Quick commerce: small “dark stores” in each neighbourhood that need frequent replenishment to deliver in minutes.
• Direct to consumer (D2C): the brand’s own website, with single-unit parcels and returns.

Each channel orders in different sizes and at different speeds, so it needs a different flow. That is the real design question: which channels share warehouses, stock and trucks, and which need their own?

Omnichannel means the customer can buy, collect or return through any channel, so the stock behind them is shared rather than split. Common patterns are ship-from-store, buy online and pick up in store (BOPIS), and one shared stock pool feeding stores and web orders. Sharing stock pools inventory and avoids one channel stocking out while another sits on excess. The catches are more complex systems, store staff picking online orders, and channel conflict, when channels compete on price or for scarce stock.

Each channel orders differently

The drivers are order size, picking unit (pallet, case or single item), delivery frequency, packaging, payment terms and the return rate. A channel strategy that ignores them can grow a channel that loses money on every order.

Sharing stock across channels

• Separate pools: each channel has its own stock. Simple and protective, but one channel can run out while another sits on excess.
• Shared pool with rules: one stock serves all channels, with allocation rules that protect key commitments (a retailer with penalty clauses, a marketplace promotion) and release the rest to whoever orders first.
• Available-to-promise by channel: the system shows each channel what it can promise, updated as stock and orders change.

Choosing where each online order ships from

When stock sits in several places (warehouses, dark stores, shops), a distributed order management system chooses the node for each order: the cheapest that can meet the promised time, while avoiding emptying any one location and splitting orders as little as possible.

Channel conflict

Channels compete for customers, for scarce stock and on price. A deep discount on a marketplace undercuts the brand’s distributors; a large retailer’s promotion soaks up stock the kirana channel needed. Managing conflict is partly commercial (pricing, assortment by channel, channel-specific packs) and partly supply-chain (allocation rules agreed in advance and visible to everyone).

Distributor economics in general trade

Distributors earn a margin on what they sell, carry stock and credit, and run delivery routes to small stores. For the brand, the distributor network is a way to reach hundreds of thousands of outlets without owning the last mile. The trade-offs to manage: distributor stock levels (too high ties up their cash, too low loses sales), the frequency of visits to each outlet, and the data the brand gets back. Distributor management systems give brands visibility of secondary sales, from distributor to retailer, which is far closer to real demand than primary shipments to distributors.

Quick commerce

Dark stores hold a few thousand items close to dense neighbourhoods and are replenished frequently, sometimes several times a day, from a mother warehouse. For brands, this means smaller, more frequent orders, tighter delivery windows, and forecasting at very local level. Stock in each dark store is small, so availability depends on the speed and reliability of replenishment rather than on large buffers.

Ship-from-store and pickup

Using shop stock for online orders pools inventory across channels and cuts delivery distance, but it needs accurate store inventory records (an online order for an item that isn’t really on the shelf becomes a cancellation), staff time for picking and packing, and rules about which stock stays for walk-in customers. Research on retailers that showed shoppers real-time store availability online and offered in-store pickup (Gallino and Moreno, 2014) found the channels interact: online information changed where and how customers bought, so channel results can’t be judged in isolation.

Shared or separate buildings?

Sharing a warehouse across channels saves rent and pools stock, but the processes conflict: full-pallet retail orders, case-pick distributor orders and single-item parcels need different storage, picking methods and packing. A common answer is one building with separate zones and processes; a dedicated e-commerce site makes sense once parcel volumes are large enough to justify automation built for single items.