What is Weighted average cost?

How to calculate weighted average cost

The weighted average cost formula is: total cost of goods available for sale divided by total units available for sale. 'Available for sale' means everything you owned during the period, beginning inventory plus every purchase, valued at landed cost. The result is the average cost per unit, which you then apply to both the units sold (COGS) and the units left on hand (ending inventory).

Worked example: you start with 100 units that cost $10 each ($1,000), then buy 200 more at $13 each ($2,600). Total cost available is $3,600 across 300 units, so your weighted average cost is $3,600 / 300 = $12 per unit. Sell 250 units and your COGS is 250 x $12 = $3,000; the remaining 50 units are valued at 50 x $12 = $600. Every unit carries the same blended $12 regardless of which purchase order it physically came from.

Notice what the math does. The new purchase was twice the size of your opening stock, so the blended cost sits closer to $13 than $10. A large receipt at a new price moves the average hard; a tiny one barely nudges it. Eyeball a simple average of $10 and $13, write down $11.50, and you're off by fifty cents a unit.

• •WAC per unit = total cost of goods available for sale / total units available for sale
• •Beginning: 100 units x $10 = $1,000
• •Purchase: 200 units x $13 = $2,600
• •Available: $3,600 / 300 units = $12 per unit
• •Sell 250 units: COGS = 250 x $12 = $3,000; ending inventory = 50 x $12 = $600

Perpetual vs periodic weighted average

There are two flavors of WAC. Periodic weighted average computes one blended cost at the end of the period using all purchases for that period, which is simple but means you cannot value a sale until the period closes. Perpetual weighted average recalculates the blended cost on every single receipt, so at any moment you have an up-to-date unit cost and an accurate inventory value, which is what high-velocity ecommerce sellers actually need.

The perpetual method is more work by hand but far more useful, because it lets you book accurate COGS on each sale in real time rather than waiting for a month-end true-up. This is exactly the kind of calculation accounting software should automate. BeanHawk computes perpetual WAC from landed cost, recalculating the blended unit cost on every inbound receipt, so your inventory valuation and COGS stay correct between closes without manual journal entries.

The two methods produce different numbers. Say the 250 units didn't all sell at once: 80 before the second purchase order arrived, 170 after. Perpetual WAC charges the first 80 at $10 ($800) and the remaining 170 at the recalculated $12 ($2,040), so COGS is $2,840. Periodic charges all 250 at $12 and reports $3,000. Same purchases, same sales, a $160 gap on one SKU. Neither is wrong, but you can't switch between them month to month and expect your margin trend to mean anything.

Weighted average cost vs FIFO

WAC and FIFO are the two methods most ecommerce sellers choose between (LIFO is rarely used and is barred under IFRS). FIFO assumes the oldest cost layers are consumed first, so in a period of rising costs FIFO leaves your most recent, higher costs in ending inventory and reports lower COGS, which means higher reported profit. WAC blends everything together, so it lands between the extremes and reacts more gently to price swings.

Neither is 'more correct'; they are different conventions, and your books should pick one and apply it consistently. WAC wins on simplicity and is ideal when you buy the same SKU repeatedly and do not care about tracing individual batches. FIFO wins when you need cost layers to reflect physical flow, for instance when shelf-life or batch tracing matters. The practical takeaway: in periods of rising unit costs, WAC typically reports a higher COGS and lower profit than FIFO, and you should know which way your method leans.

Run the same numbers through FIFO. Selling 250 of those 300 units consumes all 100 opening units at $10 ($1,000) plus 150 newer units at $13 ($1,950), giving COGS of $2,950 and ending inventory of $650. WAC gave COGS of $3,000 and ending inventory of $600. FIFO reported $50 more profit. Switching between the two later is an accounting change, not a settings toggle, so talk to your CPA first.

Why landed cost is the right input for WAC

A weighted average cost is only as accurate as the unit cost you feed it. If you average only the supplier's invoice price and ignore freight, duty, and prep, your WAC understates true cost and your gross margin looks better than it is. The correct input is landed cost, the all-in per-unit cost once every inbound expense is allocated, so the blended figure reflects what each unit actually cost to get into Amazon's warehouse.

Getting this right matters because COGS flows straight to your profit and loss statement and your taxable income. A WAC built on landed cost gives you a defensible inventory value, an accurate gross margin, and books that reconcile to cash. BeanHawk rolls freight, duty, and prep into landed cost before computing the blended average, so the cost of goods sold on every sale reflects reality rather than just the supplier's invoice.

Timing is the hard part. Your freight invoice often arrives weeks after the container clears and the units are already selling. Compute WAC on invoice price at receipt, post the freight bill to a generic shipping expense account, and the freight never reaches inventory: your blended cost stays permanently light. Accrue estimated freight at receipt, recalculate when the real invoice lands, adjust the remaining on-hand units. Sellers doing amazon fba accounting at scale build this true-up into the monthly close.

How weighted average cost shows up in your books

In double-entry terms, WAC governs one movement: value leaving the Inventory asset account and landing in COGS when a unit sells. Receiving stock debits Inventory and credits Accounts Payable at landed cost. Selling a unit debits COGS and credits Inventory at the current blended rate. The sale itself is a separate entry, and the two should never be combined.

This is where a lot of ecommerce accounting goes sideways. Many sellers expense every inventory purchase the moment the supplier is paid, so profit collapses in restocking months and spikes in quiet ones. That's a cash flow view, not an accrual one. WAC holds purchase cost on the balance sheet until the unit sells, then releases it in proportion to the revenue it generated.

The mechanics in a general ledger depend on your tooling. QuickBooks and Xero both handle inventory, but their native item costing was designed for a shop with dozens of SKUs, not a seller pushing thousands of FBA order lines a month. Most sellers using quickbooks for amazon sellers keep SKU-level costing in a dedicated subledger and post only a summarized monthly COGS journal to the ledger. The ledger stays fast; the detail stays queryable.

• •Receipt: debit Inventory, credit Accounts Payable, at landed cost per unit
• •Freight bill arrives later: debit Inventory, credit Accounts Payable, then recalculate WAC
• •Sale: debit COGS, credit Inventory, at the current blended unit cost
• •Shrinkage or disposal: debit an inventory adjustment expense, credit Inventory, at WAC

Common mistakes that break your weighted average

The most common error is mixing units of measure. You buy a case of 24, receive it as one unit, then sell singles, and your blended cost per unit is 24 times too high. Record every receipt in the same sellable unit as the sale.

Second: returns. When a customer returns a sellable unit, it should re-enter stock at the blended cost it left at, not at zero. Booking returns at zero cost quietly deflates your average over time and inflates margin on everything you sell afterward. Amazon's returns and removals reports are the source of truth, and reconciling them monthly is unglamorous but necessary.

Third: negative inventory. If your system records a sale for a SKU it thinks has zero units, the calculation has no denominator and either errors out or silently reuses the last known cost. This happens constantly with FBA, because inbound shipments get received into Amazon's network before you've recorded the receipt. Fix the sequencing rather than papering over it.

Fourth, and the one that costs real money: bundles and multipacks. A 3-pack sold under its own SKU needs its own cost, built from three times the component's blended cost plus prep. Assign the single-unit WAC to a multipack and you'll report a third of the true COGS and congratulate yourself on a high-margin product you don't have.

Edge cases in FBA and multichannel selling

Inventory sitting in Amazon's fulfillment centers is still yours, and it still carries your blended cost. Units Amazon loses, damages, or disposes of get written off at WAC, and any reimbursement Amazon later pays lands as separate income rather than a reduction of COGS. Netting reimbursements against inventory value feels tidy but hides how much you're actually recovering.

Multichannel adds a wrinkle. The same physical SKU might sell on Amazon, eBay, and your Shopify store, and it should carry one blended cost across all three. Track cost per channel and you'll end up with three different WACs for one product. Pool cost at the SKU level, attribute only revenue and fees per channel: that's how decent multi channel inventory management software is designed.

Currency is the last complication. Pay a supplier in a foreign currency and the landed cost feeding your WAC should use the rate on the transaction date, not today's rate. Lock the cost at receipt and let FX gains or losses sit in their own account.

Automating WAC without a spreadsheet

Honest answer first: with fewer than about twenty SKUs and a few restocks a year, a well-built spreadsheet is fine. Columns for date, quantity in, landed cost, quantity out, running units, running value, and a formula recalculating the blended rate on each receipt does the job. It fails when the row count outgrows your patience or a formula silently breaks.

Past that point, look for tooling that does three things: compute perpetual WAC from landed cost rather than supplier invoice price, handle returns, removals, and adjustments without you writing journal entries, and post a summarized COGS journal to your ledger instead of dumping thousands of lines into QuickBooks. A2X and Link My Books are the established settlement-to-ledger connectors, dedicated amazon inventory management software handles the stock side, and BeanHawk covers costing and sync together. Compare on how each values inventory, not just on how prettily it maps deposits.

One test worth running during any trial: push a real month of data through it and check whether ending inventory value ties to your own count times your blended cost.

Frequently asked questions

What is weighted average cost?

It is an inventory valuation method that pools all units of a SKU and divides total inventory cost by total units to get one blended cost per unit. That blended figure is applied to every unit sold, so you do not have to track which specific batch a sold unit came from.

How do I calculate weighted average cost?

Divide the total cost of goods available for sale (beginning inventory plus all purchases at landed cost) by the total units available for sale. For example, $3,600 across 300 units gives a weighted average cost of $12 per unit, which you apply to both COGS and ending inventory.

What is the difference between weighted average and FIFO?

FIFO consumes the oldest cost layers first; weighted average blends all costs into one figure. In a period of rising costs, FIFO reports lower COGS and higher profit, while WAC sits between the extremes and reacts more gently to price swings. Pick one and apply it consistently.

What is the difference between periodic and perpetual weighted average?

Periodic computes one blended cost at the end of the period; perpetual recalculates the blended cost on every receipt so your unit cost is always current. High-velocity ecommerce sellers generally need perpetual WAC to book accurate COGS on each sale in real time.

Should I use supplier price or landed cost for WAC?

Use landed cost, the all-in per-unit cost including freight, duty, taxes, and prep. Averaging only the supplier invoice understates true cost, inflates gross margin, and distorts taxable income. The blended average is only as accurate as the unit cost you feed into it.

How do customer returns affect weighted average cost?

A sellable return comes back into inventory at the same blended cost it left at, reversing the original COGS entry. Booking it at zero cost drags your average down and inflates margin on every later sale. Unsellable returns get written off at WAC instead of restocked.

Does QuickBooks handle weighted average cost for Amazon sellers?

QuickBooks Online supports average costing on inventory items, but it was built for a modest SKU count, not thousands of FBA order lines a month. Most sellers keep SKU-level costing in a subledger and post a summarized monthly COGS journal into QuickBooks. If you're shopping for amazon accounting software, ask whether it computes perpetual WAC from landed cost or just maps settlement deposits.

What should amazon seller accounting software do with inventory costing?

At minimum: value inventory at landed cost, recalculate the blended rate on every receipt, handle returns, removals, damage, and bundle SKUs without manual journals, and hand your ledger a balanced summary rather than raw transaction dumps. A2X and Link My Books are worth comparing on the settlement side; BeanHawk combines perpetual WAC costing with the ledger sync.

Related terms

Go deeper

• COGS for Amazon Sellers: Getting Cost of Goods Sold Right