The point of an inventory KPI is to turn a pile of stock and transactions into a decision: order more, hold less, recount, or leave it alone. A number that does not change one of those is decoration.
Inventory KPIs fall into three families. Cash metrics (turnover, days on hand, carrying cost) tell you how much money the stock ties up. Service metrics (service level, fill rate, stockout and backorder rates) tell you whether customers get what they ordered. Accuracy metrics (inventory record accuracy, shrink) tell you whether the other two can be trusted. The failure mode is reading one family in isolation and optimising it at the expense of the others.
Core inventory KPIs and their formulas
| KPI | Formula | Question it answers |
|---|---|---|
| Inventory turnover | Cost of goods sold / average inventory at cost | How many times did stock sell and get replaced? |
| Days inventory on hand | 365 / inventory turnover | How many days of sales does current stock cover? |
| Inventory to sales ratio | Inventory value / sales for the period | How much stock is held per unit of sales? |
| Carrying cost rate | Annual holding costs / average inventory value | What does it cost to keep stock for a year? |
| Inventory record accuracy | Locations counted with no variance / locations counted | Can the system figure be trusted? |
| Cycle service level | Replenishment cycles with no stockout / total cycles | How often does an item make it through a cycle without running out? |
| Fill rate | Units shipped from stock / units ordered | What share of demand was met immediately? |
| Backorder rate | Order lines backordered / total order lines | How often do customers wait for stock? |
| Sell-through rate | Units sold / units received in the period | How fast does new stock sell? |
| Aged or dead stock share | Value with no movement in a set period / total inventory value | How much cash is stuck in stock that is not selling? |
| GMROI | Gross margin / average inventory at cost | How much margin does each unit of inventory investment earn? |
Each of these has a deeper guide on this site. Start with the inventory turnover ratio, carrying cost, and service level if you are building a KPI set from scratch.
A worked example across the KPIs
Here is an example for one distributor over one year, with round numbers so the arithmetic is easy to follow.
| Input | Value |
|---|---|
| Cost of goods sold | $2,400,000 |
| Average inventory at cost | $400,000 |
| Gross margin | $1,200,000 |
| Annual holding costs (storage, capital, insurance, shrink, obsolescence) | $100,000 |
| Order lines received | 50,000 |
| Order lines backordered | 1,500 |
| Locations cycle counted | 2,000 |
| Locations with a variance | 90 |
| KPI | Calculation | Result |
|---|---|---|
| Inventory turnover | 2,400,000 / 400,000 | 6.0 turns |
| Days inventory on hand | 365 / 6.0 | About 61 days |
| Carrying cost rate | 100,000 / 400,000 | 25% a year |
| Backorder rate | 1,500 / 50,000 | 3.0% of lines |
| Inventory record accuracy | (2,000 minus 90) / 2,000 | 95.5% |
| GMROI | 1,200,000 / 400,000 | 3.0 |
In this example, reading the numbers together tells a story that no single number does. Turnover looks reasonable, but roughly 1 location in 22 has a variance, so part of the 3% backorder rate is likely phantom stock: units the system shows as available that are not on the shelf. The first action is a counting and receiving fix, not a larger safety stock. Adding stock to cover inaccurate records would lower turnover and raise carrying cost without solving the cause.
How to read inventory metrics together
The metrics trade off against each other, so read them in pairs.
- Turnover with service level. Turnover can always be raised by cutting stock. The real goal is higher turns at a held service level, so report both on the same page.
- Service level with fill rate. A high cycle service level can coexist with a poor fill rate when the stockouts that do happen are large. Our guide to service level vs fill rate explains what each number actually promises.
- Accuracy under everything. Every other KPI is calculated from the stock record. If record accuracy is weak, fix it with cycle counting before tuning anything else.
- Averages hide the tail. A healthy company-wide turnover can hide a large pool of dead stock. Split every KPI by ABC class so the fast movers do not mask the slow ones.
- Inventory metrics with fulfillment metrics. Stock KPIs end at the shelf. For the customer side, pair them with order cycle time, pick accuracy, and perfect order rate from the order fulfillment process.
What a healthy result looks like
There is no universal target. A grocery chain and an industrial spare parts distributor should have very different turnover and days on hand. Judge a KPI against three references, in this order:
- Your own trend. Is the number improving over the last 12 months, and did the change come from a real operational lever or from a one-off event?
- Your sector. Compare with published peers in the same industry and business model.
- The macro picture. For U.S. businesses, the Census Bureau reports a total business inventories to sales ratio every month. At the end of July 2026 it was 1.30, compared with 1.37 in July 2025 (Census Bureau Manufacturing and Trade Inventories and Sales, released September 16, 2026). Your own ratio will differ by sector, but the direction of the national figure helps explain supplier and demand conditions around you.
Choosing the right KPI set for your business
| Business type | Lead KPIs | Why |
|---|---|---|
| E-commerce and retail | Sell-through, fill rate, days on hand, aged stock share | Seasonal assortments and fast markdown decisions |
| Wholesale distribution | Fill rate, backorder rate, turnover by ABC class, GMROI | Service promises to trade customers and wide assortments |
| Manufacturing | Record accuracy, raw material days on hand, component stockouts | Production stops when one component is missing |
| Spare parts and service | Fill rate for critical parts, aged stock share, carrying cost | Long tail items held for service commitments |
Keep the set small. Five or six KPIs that someone reviews every week beat twenty that nobody acts on.
Common ways inventory KPIs mislead
- Using sales instead of cost of goods sold in turnover inflates the ratio by the margin.
- Using a single year-end balance instead of an average makes turnover depend on the day you counted. Use monthly averages at least.
- Measuring accuracy by value only. Offsetting overs and unders cancel out in value while individual SKUs are badly wrong. Measure at SKU and location level.
- Counting a backorder as a sale. Service metrics should use the date the customer asked for, not the date you eventually shipped.
- Ignoring the cost of holding stock. A high GMROI can hide rising storage and obsolescence cost; review carrying cost alongside it.
A simple reporting cadence
| Frequency | Review | Owner |
|---|---|---|
| Daily | Stockouts and backorders on A items, count variances found | Inventory control and warehouse |
| Weekly | Fill rate, service level, aged stock additions | Planning and purchasing |
| Monthly | Turnover, days on hand, carrying cost, GMROI, record accuracy by zone | Operations and finance |
| Quarterly | KPI targets, ABC classes, safety stock settings | Leadership with planning |
Each review should end with an action: reorder, reduce, recount, clear, or leave alone. If the same metric misses its target three reviews running, the target or the process needs to change.