Home / Blog / Inventory Turnover
How-to · 2026

Why Inventory Turnover Is Not in Your Dashboard

The formula is two lines and the denominator is not in any file your store exports. What average inventory actually requires, why a current stock level is not it, and the questions worth asking instead.

By · Published August 20, 2026 · 7 min read
Twenty-five years starting companies, consulting, owning a multi-million-dollar P&L and leading business intelligence for large enterprises.

Inventory turnover is a standard retail metric, it appears on every list of e-commerce KPIs, and most stores cannot calculate it from the data they have. Not because the tool is limited. Because half of the formula describes something a sales export does not contain.

This is the boundary, stated plainly, and the things worth measuring instead.

The formula, and where it breaks

Inventory turnover = cost of goods sold / average inventory at cost
Days of inventory  = 365 / turnover

The numerator is fine. Cost of goods sold is quantity sold multiplied by unit cost, summed over the period, and both halves of that are obtainable — order lines carry quantity, and a variant cost list carries the rest, as covered in gross margin by SKU.

The denominator is the problem, and it is a structural one rather than a data-quality one.

Balances and events are different kinds of thing

An order export is a record of events: on this date, this happened. Every figure you can build from it is a sum of events over a window, which is why sales, discounts, refunds and cost of goods all work.

Average inventory is a balance: at any moment, this much stock was sitting there. Balances are not derivable from event records unless you have every event that ever moved them — every purchase order, every receipt, every adjustment, every write-off, every count correction, going back to a known starting quantity. Sales alone are a fraction of that list.

This is the same shape as the capacity problem in professional services, where utilization needs hours that were available and a timesheet only records hours that were logged. Different industry, identical failure: the denominator is a property of the world, not of the transaction file.

A current stock level is not an average

The tempting shortcut is the stock-on-hand figure your inventory system shows today. It is one observation of a balance, taken now, and turnover needs the balance across the whole period.

For a store with flat inventory that is a reasonable approximation. For a store that received a container in October, or ran down stock into a sale, or is seasonal in any way, it is not:

Stock taken 12 Oct, week before delivery      41,000
Stock taken 26 Oct, week after delivery      198,000
Cost of goods sold, twelve months            890,000

Turnover on the first        21.7×
Turnover on the second        4.5×

Same business, same year, same formula, and a figure that differs by a factor of five depending on which Tuesday somebody ran the export. A number that unstable is not a measurement.

From twenty-five years of being asked for this

The request always arrives the same way: someone read that turnover is the metric that matters in retail, and they are not wrong. What follows is the awkward part, because the honest answer is that the data was never kept, and the person asking has usually already seen a turnover figure somewhere — produced by exactly the current-stock substitution above.

What I have learned to do is separate the two conversations. The first is what we can answer today, which is more than people expect. The second is what to start recording so that in a year the answer is yes. A monthly stock-at-cost snapshot is a five-minute export and a folder. It is the cheapest analytics investment a store can make, and it is only ever made by people who have already been told no once.

What your export can actually answer

Four questions, all answerable from order lines plus a cost list, none of which require a balance:

And one more, if you have a current stock level and are willing to state the assumption: days of cover, current units divided by recent daily sales rate. It assumes the recent rate continues, which is exactly as true as it sounds, but it uses today stock level for something today stock level can legitimately support — a forward projection rather than a historical average.

How to make turnover available next year

Export stock on hand at cost, per SKU, once a month, and keep the files. That is the whole requirement. Twelve of those and average inventory becomes a real calculation instead of a guess.

Most inventory systems will do this on a schedule. The reason almost no store has the history is that nobody asks for it until the day they want turnover, and by then the previous year cannot be recreated — a balance that was not recorded is gone.

What we will and will not compute

Quiriz for e-commerce computes net sales, gross profit and contribution margin from the store, refund and ad exports, broken down by product category or channel.

Inventory turnover is not on that list, and asking for it returns a decline with the reason rather than a number. That is deliberate. Every calculation the product could offer for turnover would require substituting a current stock level for an average, and a metric that changes by a factor of five depending on the export date is worse than no metric, because it will be believed.

If you have kept monthly stock snapshots, that changes — the data exists, it can be uploaded, and the calculation becomes real. Without them, the honest answer is the one above.

If you take one thing from this: start exporting stock on hand at cost once a month and keep the files. It costs five minutes, and it is the only way the metric you want becomes available at all.

Ask the questions your export can actually answer

Upload your order lines and a cost list, and ask for margin and revenue by product in plain English. Free to start.

Try Quiriz free →

Frequently asked questions

What is the inventory turnover formula?
Cost of goods sold for a period divided by average inventory at cost over that same period. Some businesses use sales instead of cost of goods sold in the numerator, which produces a higher number that is not comparable to anyone else. Both halves must be at cost for the ratio to mean what it is normally taken to mean.
Why can I not calculate inventory turnover from my Shopify export?
The numerator is available — quantity sold times unit cost is in the order lines. The denominator is not. Average inventory is a balance held over time, and an order export records events, not balances. Nothing in a file of sales says how much stock was sitting in the warehouse in March.
Can I use my current stock level as average inventory?
Only if stock has been genuinely flat, which for most stores it has not. A current level is one observation taken today. Substituting it for a twelve-month average produces a turnover figure that reads as authoritative and depends entirely on when you happened to run the export — one taken the week before a delivery and one taken the week after will disagree substantially.
What can I calculate from an order export instead?
Units sold and revenue per SKU, sell-through against a starting quantity if you know it, gross margin per SKU with a cost list, and revenue concentration. Days of cover is also available if you have a current stock level, on the stated assumption that recent sales rate continues. All of those are honest answers built on data that exists.
What would I need to track to get real turnover?
Periodic inventory snapshots: stock on hand at cost, per SKU, recorded at a regular interval and kept. Monthly is usually enough. Most inventory systems can produce this and most stores do not retain it, because nobody asked until the day someone wanted turnover — at which point the missing history cannot be recreated. Start keeping snapshots now and the metric becomes available in a year.