Inventory Days on Hand: How to Calculate and What It Tells You

Days on hand is current inventory divided by average daily demand. A thousand units selling at twenty-five a day is forty days on hand. The number translates stock into time, which is the unit purchasing decisions are actually made in.

What is days on hand really measuring?

It converts a quantity into a deadline. Telling a founder you hold five thousand units communicates very little. Telling them you hold forty days of cover, against a supplier who takes twenty-one days to deliver, tells them they have about nineteen days to place the order.

That is the whole value of the metric, and it is worth being precise about what it is not:

• It is not a performance score. A high days on hand is not automatically waste and a low one is not automatically lean. Both are only readable against your lead time.
• It is not a forecast. It extrapolates today's stock at one assumed rate of sale. Change the rate and the answer changes, which is why the choice of rate is the real decision.
• It is not inventory turnover, though it is the same information. Turnover is the annualised, dollar-denominated view that finance uses; days on hand is the per-SKU view that purchasing uses.

What is the days on hand formula?

Days on hand = current inventory ÷ average daily demand

Both terms need a definition before the division means anything. Current inventory is normally sellable stock in your own warehouse or 3PL, excluding anything on the water. Average daily demand is where the three methods below differ, and the difference is not cosmetic — on a seasonal SKU they can be a fortnight apart.

DOH, DSI, DIO: which name are you looking for?

Four terms, two actual metrics. The vocabulary is split roughly along the line between operations and finance, and mixing them is the most common way this calculation goes wrong.

The rule that follows is short: use the unit version to buy, use the dollar version to report, and never quote one where the other is expected. A DSI of forty days across a catalogue can sit on top of a hero SKU at twelve days and a discontinued flavour at four hundred, and the average conceals both problems. The dollar-level relationship to turns is covered in inventory turnover.

How do you calculate it? Three methods, and when each is right

Method 1: historical average

DOH = current inventory ÷ (sales over a recent period ÷ days in that period)

For example, say you hold 1,000 units and sold 750 in the last 30 days. Average daily demand is 25 units, so days on hand is 40. It uses real sales, it takes a minute, and it assumes next month looks like last month.

Use it for stable SKUs and for a first pass across a whole catalogue. Its failure mode is predictable and one-directional: going into a peak it reads high, coming out of one it reads low, and in both cases it is wrong exactly when the decision matters most.

Method 2: forward-looking forecast

DOH = current inventory ÷ forecast daily demand

For example, suppose the same 1,000 units face a forecast of 900 over the next 30 days. Forecast daily demand is 30, so days on hand is 33 — a week shorter than the historical method said, on the same stock. This is the version that should drive a purchase order, because a purchase order is a bet on the future and not on the past. It is only as good as the demand forecast underneath it.

Method 3: the cost-of-goods method

DSI = (average inventory value ÷ COGS) × days in the period

For example, imagine average inventory of $50,000 against annual COGS of $450,000. That gives roughly 41 days. This is the accounting version, it works in dollars, and it describes the business rather than the product. Report with it; do not order with it.

What is a good days on hand?

This page does not give you a benchmark, and the omission is deliberate.

Published days-on-hand ranges by category are easy to find and almost all of them trace back to each other rather than to a measurement. More importantly, even an accurate industry figure would not be actionable, because the correct answer is determined by two things that are specific to you and both of which you already know.

Your floor is your lead time plus your safety stock, expressed in days.

• Average lead time in days, taken from your purchase order receipt dates rather than from the supplier's quote. The gap between those two is usually material.
• Plus your safety stock in days of cover. If you size safety stock in units, divide it by average daily demand to convert.
• That sum is the level at which you should already have ordered. Not the level at which you start thinking about it.

Say your receipts show an average lead time of twenty-one days and you hold a week of safety stock. Your floor is twenty-eight days. At thirty days on hand you are inside the window; at twenty-five you are late and the only remaining questions are about mitigation. The arithmetic is the same one behind the reorder point, expressed in time instead of units.

Three things move that floor, and they are worth reviewing quarterly rather than annually:

• Lead-time variability. A supplier who ranges widely around their average forces a higher floor than their average alone suggests. The measurement lives in safety stock.
• Shelf life. On perishable SKUs the ceiling can bind before the floor does. Days on hand approaching remaining shelf life is a markdown decision, not a purchasing one.
• Cash. A floor you cannot fund is still the floor. It just means the conversation is about order frequency or supplier terms instead of quantity.

What do you do the week it drops below the floor?

The metric has done its job at this point. What follows is a decision, and it is usually made badly under time pressure, so it is worth having the order of operations written down.

1. Check the denominator before the numerator. A days-on-hand figure that fell sharply is often a demand spike that has already passed, or a stockout week that depressed the historical average and is now correcting. Confirm the sales rate is real before committing cash to it.
2. Recalculate on the forecast, not on history. If method 1 and method 2 disagree, method 2 is the one the purchase order has to answer to.
3. Decide whether you are buying cover or buying a launch. Topping up to the floor and buying for a promotion are different orders with different risk, and merging them is how brands end up with the wrong quantity of the right product.
4. If the order cannot arrive in time, stop optimising the order. Partial shipments, air freight on the fast-moving portion, and telling your largest account early are all worth more than a perfectly sized purchase order that lands two weeks after the shelf empties. What that shortfall actually costs is covered in fill rate.
5. Record why. The reason this SKU went low is the input to next quarter's safety stock, and it is the one input nobody writes down.

How Planster uses days of cover

Days on hand is a number you compute. What you actually want is the decision on the other side of it, which is whether to place an order this week and for how much.

Planster pulls sales and inventory from 150+ systems, builds one demand forecast across DTC, Amazon, retail and wholesale, and nets it against what is on hand, on order and already committed. From that it computes each SKU's upcoming orders using starting inventory, target levels, safety stock and lead times, and ranks them by when they are due — so the output is the list of what has to ship this week rather than a column of days-of-cover figures to sort yourself. It flags the other end too: SKUs carrying excess months of supply, which is the dead stock problem arriving early enough to do something about. Every order still needs you to approve it.

The honest limit: none of that is a reason to stop understanding the formula. A brand with a few dozen SKUs and one reliable supplier can run days on hand in a spreadsheet and be entirely well-served. The work only becomes unmanageable when the number of SKUs multiplied by the number of supply routes exceeds what anyone will actually re-check on a Monday — and the tell is not a stockout, it is that the spreadsheet's lead times were last updated by someone who has left. Planster's full app, Plan, is $999/month. The wider decision it sits inside is supply planning.