Inventory Accounting for Online Sellers 2026

When a seller sends stock into a marketplace fulfilment programme, the goods do not change hands. The platform holds them as a service provider. The seller continues to own the inventory and bears the risk on it until the sale to the end customer.

It follows that the stock belongs on the seller's balance sheet. Treating the despatch to the fulfilment centre as though it were a sale, or simply losing sight of the stock once it leaves the godown, overstates cost of goods sold in that period and understates closing inventory.

Every month, three numbers should agree for each SKU: units in your own storage, units the platform reports as held at its centres, and units in the stock ledger.

Where they do not, the difference is almost always one of a small number of causes, each of which is a genuine accounting event rather than a rounding issue.

A blended cost of goods sold figure is not much use to a multi-channel seller, because the same unit can carry different fulfilment and storage costs depending on where it was held and how it shipped.

The workable approach is to keep product cost consistent across channels, then attribute fulfilment, storage and logistics separately by channel. Product cost tells you the floor. Channel attribution tells you which route to market is actually worth serving.

A seller who manufactures the product cannot take cost from a purchase invoice, because there is not one. Material cost, conversion cost and a reasonable absorption of production overhead have to feed a cost per unit before any margin figure means anything.

This is common in Pune, Chakan and Ahmedabad's textile belt, where manufacturer-sellers routinely list their own output. Skipping the costing step makes the marketplace P&L look considerably healthier than it is, because the entire manufacturing overhead sits below the line instead of in cost of goods sold.

Yes. Sending stock into a fulfilment programme is not a sale, and the seller retains ownership and risk until the goods are sold to the end customer. It must therefore remain in the stock ledger at cost and appear in closing inventory on the balance sheet.

Monthly at minimum, matching your own counts, the platform's inventory reports and the stock ledger for each SKU. Sellers with high volumes across several fulfilment centres often move to a fortnightly cycle, because the volume of movements makes monthly differences harder to trace back to a cause.

The common causes are stock in transit, returns despatched but not yet received back, units lost or damaged in the platform's custody awaiting reimbursement, and SKU mapping errors where one product carries different codes on different channels. Each is a real event that belongs in the books.

Either is acceptable if applied consistently, and weighted average is generally simpler for high-volume sellers with frequent replenishment at varying prices. What matters more than the choice is consistency, since switching methods mid-year distorts comparability and invites questions at audit.

Build cost per unit from material, conversion cost and a reasonable absorption of production overhead, rather than from any purchase document. Manufacturer-sellers who skip this step and use only material cost systematically overstate marketplace margin, because the whole of manufacturing overhead falls below the gross margin line.