The inventory shrinkage formula, with real numbers

Shrinkage is not a mystery metric; it is subtraction. Your records say you own a certain amount of stock, a count says you own less, and the difference — valued in money — is what shrank. The formula is one line. The hard part is producing the two numbers honestly and then reading the result without lying to yourself about the cause. This page keeps the arithmetic on the surface: every term defined, a worked example from recorded value to margin impact, the classes of cause behind the gap, and the exact path from a Shopify count to a shrinkage figure you can defend.

##The two terms

Recorded inventory value

What your system believes you own, valued in money: on-hand quantity × unit cost, summed across the variants you are measuring. In Shopify this is the on-hand quantity at a location multiplied by the variant's cost per item. Value it at cost, not retail — shrinkage is a loss of the asset you paid for, and mixing in retail markup inflates the number and makes it incomparable period to period. (The cost-vs-retail question is common enough that it has its own answer in the FAQ below.)

Counted inventory value

The same variants, same cost, but valued at the quantity a physical count actually found. If the count is clean — every unit scanned once, nothing double-counted, drafts and archived variants excluded — this is the true asset on your shelves. A sloppy count produces fake shrinkage, so the count discipline (blind scanning, one pass, reconciling anything that moved mid-count) is what makes the formula trustworthy.

##Worked example, end to end

Take one variant to keep the numbers legible. Your records show 100 units on hand, and the variant's cost is $12. A physical count finds 95 units. Five units are unaccounted for.

The $60 is a real expense, not a rounding note. Those five units were an asset you paid $60 for; the count says they are gone, so $60 leaves your inventory account and lands in cost of goods sold as shrink. That directly reduces gross margin for the period — you already spent the cash to buy stock you can no longer sell. If those units would have retailed at, say, $30 each, the *opportunity* lost is larger ($150 of potential sales), but the number you book and track is the $60 at cost. Keep the two ideas separate: cost is what shrank on the books; retail is what you also failed to earn.

Scale the same subtraction across the whole catalog and you get a store-level shrinkage rate. The mechanics do not change — recorded value and counted value are just summed across every variant before you divide.

##What the gap is actually made of

A shrinkage percentage tells you how much vanished, never why. The gap is a mix of distinct causes, and the whole point of measuring is to attribute it so you can act. The common classes:

• External theft — shoplifting, walk-outs, stock taken by people outside the business.
• Internal theft — stock taken by staff, or unrecorded 'personal use'.
• Damage and breakage — units destroyed, spoiled, or made unsellable and thrown away without being adjusted out.
• Administrative and clerical error — miskeyed receiving quantities, unrecorded sales, double-counted units, wrong-variant scans, unit-of-measure mistakes.
• Supplier and receiving fraud or error — cartons invoiced as full but shipped short, so stock you were charged for never physically arrived.

##Measuring it in Shopify

Shopify does not print a 'shrinkage %' for you, but it holds everything the formula needs. The path is: run a count, let the count value the variance at cost, then record the adjustment against the reason that names the cause — so next quarter's number can be attributed instead of shrugged at.

That third step is where most shrinkage tracking quietly fails. Shopify offers seven adjustment reasons — Correction, Count, Received, Return restock, Damaged, Theft or loss, and Promotion or donation — and the reason you pick is the only place the *cause* is ever recorded. Map the cause to the reason deliberately:

##Bringing the number down

Measurement is diagnosis; these are the levers that move the number:

• Count on a cadence, not once a year. Frequent cycle counts catch a gap while its cause is still fresh and reversible; an annual count discovers a year of blended losses with no trail back to the cause.
• Tighten receiving. A large share of 'shrinkage' is stock that never arrived. Count cartons in against the packing slip at the door, not weeks later, so short shipments become supplier claims instead of silent losses.
• Attribute every adjustment. A count is only as useful as the reason attached to it. Consistent, honest reasons turn a raw variance into a story you can act on.
• Assign accountability by counter. When each counted line records who scanned it, a persistent gap in one section stops being anonymous and becomes a conversation.

##Doing this with Solvi Stocktake

The formula is trivial; producing the two honest numbers across thousands of variants is not. Solvi Stocktake runs the count — barcode, SKU, or phone camera, with blind scanning so a bip just increments and moves on — and captures the on-hand at the moment each line is scanned, so units that sell mid-count get flagged to re-verify instead of silently corrupting the variance. When the count is done, it shows the discrepancy report with inventory valued at cost before and after on the Growth plan, so the shrinkage figure is computed for you rather than exported to a spreadsheet. Nothing is written to Shopify until you review and apply, and the apply is reversible in one click — so a bad count never becomes a bad shrinkage number you have to unwind. Free covers two counts a month; scheduled cadence reminders and cost valuation live on Growth.

Same arithmetic as the worked example above — recorded value minus counted value, over recorded value — computed continuously, attributed by reason, and safe to apply.