A brand that sells through its own storefront, two marketplaces, and a handful of countries doesn’t have a single demand pattern. It has several, and they all draw on the same stock. Put too little inventory in the wrong place and orders get delayed or cancelled. Put too much in the wrong place and cash sits idle. Multichannel growth turns stock placement from a warehouse chore into a planning decision that has to be made again and again.
Inventory allocation and the decisions it covers
Inventory allocation is the planning decision made before any order arrives. It settles which SKUs to hold, in what quantities, and which markets and warehouse locations they should sit in. In practice, inventory allocation for ecommerce means answering a few questions for every product: where demand will come from, how much stock covers it, and what it costs to hold. The answers determine what is available when an order lands, making allocation the foundation for delivery speed, availability, and margin. It is not a one-time setup. It gets revisited as sales patterns change.
Inputs for SKU-level allocation
Allocation works best when you decide it SKU by SKU, using a handful of inputs. Channel-level sales history shows where each product actually sells, since an item that moves quickly on a marketplace may crawl on the brand’s own site. SKU velocity, the pace of sales over a set period, ranks products against each other. Demand variability, meaning how much sales swing from week to week, shows how much buffer a product needs.
Margin indicates how much holding cost a product can carry, while product size and storage cost show how expensive it is to keep. Lead time, the gap between ordering stock and having it ready to ship, sets how far ahead decisions must be made. Target service level is the share of orders the brand aims to fill without a stockout. A higher target needs more stock.
A practical product-segmentation framework
Sorting products into five groups turns those inputs into a working rule of thumb.
- Fast movers: High, steady sales across channels. Hold deeper stock close to the biggest markets and replenish often.
- Stable mid-volume items: Predictable but moderate sales. Hold them in fewer locations with modest stock.
- New products: No reliable history. Start with a small quantity in one shared location, then add depth as channel-level sales appear.
- Seasonal goods: Sales cluster in a short window. Position stock ahead of the peak in the markets where the item sells, and plan how to sell down what remains.
- Long-tail SKUs: Many items with few sales each. Keep them in one central location rather than copying them across warehouses.
Hypothetical example: a brand sells a water bottle, a limited-edition colour, and a replacement lid. The bottle sells daily on every channel, so it gets stock in each major region. The new colour has two weeks of sales history and stays in one shared location. The lid sells a few units a month, so a single central bin serves all channels.
Shared inventory, safety stock and replenishment
Shared inventory means one pool of stock serves several channels instead of each channel getting its own allocation. It lowers the chance that a product sells out in one channel while sitting unused in another, as long as stock counts update fast enough to prevent overselling. Sharing works within a warehouse or region. Channels that ship from different regions still need a separate allocation for each region. Safety stock is the extra buffer held in case demand runs above forecast or a delivery arrives late. How much is enough depends on demand variability, lead time and the target service level. A replenishment trigger, often called a reorder point, is the stock level that prompts a new order. It should cover expected demand during lead time plus safety stock.
Fragmentation is the risk in the other direction. Spread a product across too many warehouses and each location holds a thin layer; each needs its own safety stock, and a stockout in one place can coexist with surplus in another. Fewer, better-chosen locations often work better than many small ones.
Inventory allocation versus order routing
The two are different decisions made at different times. Allocation happens in advance and places the stock. Order routing is the later execution step: when an order arrives, it picks the location that should fulfil it, based on available stock, destination, cost and service level. Routing works only with inventory already in the network, so it cannot repair a weak allocation. If a product was never placed near a market, routing cannot bring it closer. The sequence runs from demand data to allocation, then customer order, then routing, then fulfilment. That holds whether the network belongs to the brand or to a provider such as NextSmartShip, which offers both China-based and local fulfilment.
Keep allocation responsive
Channel demand shifts, so review allocation regularly, more often for fast movers than for the long tail. Compare what each channel sold with what each location held, then move stock or adjust reorders to match. The goal stays constant: protect availability, delivery performance, margin and working capital at once, without copying every SKU everywhere.