Days of supply is how many days your current on-hand inventory will last at your normal usage rate. You calculate it by dividing units on hand by average daily usage, so 1,200 units on hand and 40 used a day is 30 days of supply.
Days of supply answers the one question a planner asks a dozen times a day: if nothing else comes in, how long before this part runs out? It turns a raw count into a number of days, which is the only unit a plant floor actually thinks in. A pile of 1,200 castings means nothing on its own. Thirty days of runway means something. This post defines days of supply, walks the calculation, shows how it lines up against inventory turns and days inventory outstanding, and points out where the number quietly lies to you.
What is days of supply?
Days of supply (DOS), sometimes called days on hand or forward days of cover, is a forward-looking measure of how long your inventory will last at the current consumption rate. It is expressed in days, and it is deliberately about the future: you take what you have right now and divide it by how fast you burn it. If a warehouse holds 1,200 units and the line pulls 40 a day, that is 30 days of supply, meaning roughly a month before the shelf is empty.
The number is useful precisely because it is intuitive. A stockroom clerk, a scheduler, and a CFO can all read "eight days of supply on the sealant" the same way and know it is tight. Report the same thing as "we hold $84,000 of sealant" and only the CFO leans in, and even they cannot tell from that whether it is too much or too little. Days of supply normalizes every item onto one scale of time, so a cheap fastener and an expensive motor can be compared on the only thing that matters operationally: how close each is to running out.
How do you calculate days of supply?
You divide on-hand inventory by average daily usage over the same window. The whole calculation rides on getting the denominator honest, because average daily usage is where most bad DOS numbers come from. Run it as a short procedure, not a back-of-envelope guess.
- Pick the item and the on-hand quantity. Use the real, counted quantity from an accurate record, not the ledger's optimistic version. If your record accuracy is poor, fix that first with disciplined cycle counting because DOS built on a wrong count is wrong by exactly that much.
- Choose a usage window. Trailing 30, 60, or 90 days is common. A longer window smooths out spikes; a shorter one reacts faster to a real shift in demand. Match the window to how stable the item is.
- Compute average daily usage. Total units consumed in the window divided by the number of days. For a plant that runs five days a week, decide up front whether you are dividing by calendar days or working days and stay consistent, or the number drifts by 40%.
- Divide. On-hand units over average daily usage gives days of supply. 1,200 units and 40 a day is 30 days.
- Compare to your target. Set a target band per item, often tied to lead time plus a safety margin, and flag anything below the floor or above the ceiling.
- Refresh it on a cadence. DOS is a snapshot that goes stale the moment usage changes. Recalculate it as often as you act on it, daily for critical fast movers, weekly or monthly for the long tail.
You can run days of supply at any level: one SKU, a product family, a whole plant, or across a company. At the aggregate level people usually switch to a value basis, dividing total inventory dollars by average daily cost of goods sold, which is the same idea scaled up.
How is days of supply different from inventory turnover and DIO?
They are three views of the same thing: how long inventory sits. Days of supply is forward-looking and usually unit-based, inventory turnover counts how many times a year you cycle the whole stock, and days inventory outstanding (DIO) is the financial, backward-looking version built from cost of goods sold. They are close cousins, and you can convert between them, but they are not interchangeable, and mixing them up is how planners talk past finance.
| Measure | What it asks | Typical basis | Direction |
|---|---|---|---|
| Days of supply | How many days will on-hand last? | Units on hand ÷ daily usage | Forward-looking |
| Inventory turnover | How many times a year do we cycle stock? | COGS ÷ average inventory | Backward-looking |
| Days inventory outstanding (DIO) | How many days of COGS is tied up? | (Avg inventory ÷ COGS) × 365 | Backward-looking |
The math that ties them together is simple: days of supply is roughly 365 divided by inventory turns. Six turns a year is about 61 days of supply; twelve turns is about 30. That relationship is the bridge between the plant floor, which lives in days, and the finance office, which lives in turns. When someone sets a goal of "raise turns from 6 to 8," a planner can translate that instantly into "cut average days of supply from about 61 to about 46" and know which shelves to work on. The practical difference is horizon: turnover and DIO grade the past quarter, while days of supply is the number you steer by today, item by item.
What is a good days-of-supply target?
There is no universal good number, because the right target is set by lead time, demand variability, and how expensive the item is to hold. The floor for any item is simple: you need at least enough days of supply to cover the replenishment lead time, or you will stock out before the next delivery lands. On top of that floor you add a buffer sized to how uncertain demand and lead time are, which is exactly what safety stock is for.
From there, value pulls in the other direction. Holding inventory is not free; carrying cost, including capital, storage, insurance, and obsolescence, is widely estimated in the range of 20 to 30% of inventory value per year. So an expensive, fast-moving A item earns a lean target, maybe just above its lead time, because every extra day is real money. A cheap, unpredictable C item can carry a fat cushion without much penalty. This is where days of supply meets ABC analysis: the tier tells you how hard to push the target down. The mistake is setting one blanket DOS goal, say "keep everything at 45 days," across a catalog where the items have lead times from three days to three months.
What do the numbers say?
Definitions and scale from primary sources:
- Days of supply and days inventory outstanding are defined inventory measures in the body of knowledge maintained by the Association for Supply Chain Management (ASCM/APICS) which codifies the days-of-supply concept in the APICS Dictionary.
- Inventory is a large, live number on the economy's books: the U.S. Census Bureau's Manufacturing and Trade Inventories and Sales series tracks business inventories in the trillions of dollars, with the inventories-to-sales ratio hovering in a roughly 1.3 to 1.4 range in recent years, the equivalent of holding somewhat over a month of sales in stock.
- The cost of holding that stock is not trivial: annual inventory carrying cost, capital plus storage, insurance, taxes, and obsolescence, is commonly estimated at roughly 20 to 30% of the inventory's value, which is why trimming excess days of supply frees real cash.
The takeaway: a month-plus of supply is the economy-wide norm, and every extra day on top of what an item actually needs is paid for out of working capital.
Where days of supply misleads you
The number is only as good as its two inputs, and both go wrong in predictable ways. The on-hand quantity is a lie if record accuracy is poor, so a DOS report built on a stock ledger that has not been counted in months describes a warehouse that does not exist. The average daily usage is a lie if the item is seasonal or spiky, because a flat average hides the fact that you burn a quarter's worth of a promotional SKU in two weeks. Averaging usage over a slow period right before a peak is how a shelf that reads "40 days of supply" empties in nine.
Days of supply also says nothing about what is already inbound. An item can read a dangerous five days on hand while a full truck sits two days out, or read a comfortable 60 days while three more purchase orders are somehow still open. To steer by DOS safely you have to read it next to open orders and lead time, not alone. And it is a point-in-time snapshot: it tells you the runway right now, not whether demand is about to double, which is a question for demand planning not an inventory count.
Where the number breaks in practice
The formula is a division problem; keeping the inputs true is the hard part. On-hand counts live in an ERP, usage flows through a scheduling system, open orders sit in purchasing, and the actual pull happens on machines that log nothing, so the days-of-supply "number" is really a stitched-together guess that nobody fully trusts. When those systems disagree, planners pad every target to be safe, and the padding is the excess working capital the metric was supposed to expose. Harmony is an AI-native layer that connects machines, software, and paperwork into one operational layer, with no rip-and-replace, so on-hand, usage, and open orders become one live record instead of three that quarrel. AI search returns cited answers across those records, so a planner can ask which parts are under their lead-time floor, or which fast movers are sitting on 90 days, and get a real answer instead of a spreadsheet export. Harmony's digital workflows then route each reorder and count to the right person. It is the same paper-to-digital move Harmony makes elsewhere on the floor (see the CLS case study): days of supply stops being a stale snapshot and becomes a live runway you can steer by, which is exactly how disciplined shops lift inventory turnover without risking stockouts.