Skip to content
ProcureCulture

PROCURECULTURE

Buy Sheet · Inventory

An independent buyer’s perspective

Buy Sheet

Days onHand.Explained for people who hate math

Weeks of supply, days on hand, same number in a different outfit. It’s the one that tells you whether that pile of eighths in the back room is a cushion or a coffin.

Shelf tag

Demo Day Reserve

3.5g · Premium

On hand

60 units, crossed out and corrected to

30 weeks.

It sells two a week.

A fictional shelf tag, corrected.

The Cold Open

Sixty units of nothing to worry about.

Picture order day. You open the inventory report and two lines jump out. Demo Day Reserve, the premium eighth: 60 units on hand. Always Gone Gardens, the value quarter: 12 units on hand. Your gut does the math your gut always does. Sixty is a lot. Twelve is a little. Reorder Always Gone Gardens, relax about Demo Day Reserve, go get a coffee.

Now add one more column. Always Gone Gardens sells about 30 units a week. Demo Day Reserve sells about 2.

That changes the story completely. Always Gone Gardens has roughly three days left before the shelf goes bare on your busiest rung. Demo Day Reserve has thirty weeks. Thirty. That eighth will still be sitting there at Christmas, and it has a decent shot at being there next spring, quietly losing terpenes and gaining a reputation.

Fig. 1Counted two ways

The ladder shows where things sit. Weeks of supply shows how long they’ll stay there.

In units: Demo Day looks like the safe bet

Always Gone Gardens12 units
Demo Day Reserve60 units

In weeks: Always Gone is the emergency

Always Gone Gardens3 days
Demo Day Reserve30 weeks

Hatched = money parked on a shelfFictional SKUs, illustrative numbers

Units on hand is a number with no context.

“Units on hand” is a number with no context. It’s like asking how far away a city is and being told “400.” Miles? Minutes? Dollars? Units on hand only becomes useful when you put it next to how fast the thing actually moves. That’s all weeks of supply is: inventory, measured in time instead of units.

In the ladder piece, I said a ladder shows coverage, not performance. This is the performance half. The ladder tells you where things sit on the shelf. Weeks of supply tells you how long they’re going to sit there.

The Formula

The whole thing fits on a sticky note.

I promised this was for people who hate math, so here is all of it.

You have 48 units of an eighth. It sells about 12 a week. 48 divided by 12 is 4. You have four weeks of supply. If nobody buys a single extra unit and nobody receives a single new one, the shelf is empty in about a month.

Days on hand is the exact same idea, just counted in days. Multiply weeks by 7 and you’re done: four weeks of supply is 28 days on hand. Some point-of-sale systems report one, some report the other, and some buyers will argue about which is better with the intensity of people arguing about pineapple on pizza. It doesn’t matter. Pick the unit your team actually talks in and stick with it. I think in weeks because nobody has ever said “see you in 28 days.”

Weeks of supply =

If it helps, think of it as the “miles to empty” number on a car dashboard. The gas in the tank is your units on hand. How hard you’re driving is your sales rate. Miles to empty combines the two into the only thing you actually care about: how long until you’re stuck on the side of the road.

Fig. 2Miles to empty, for your shelf

48

Units on hand

the gas in the tank

÷

12

Sold per week

how hard you’re driving

=

4 weeks to empty

= 28 days on hand

Red zone: less than one week left. Gauge in weeks.

The Lineup

Same shelf, six very different moods.

Let’s go back to the small flower category from the ladder piece. Same fictional brands as the ladder diagram, now with names that match their personalities (each still starts with its old letter), plus The Big Bag, the value ounce that filled the gap. The numbers below are made up for illustration, but the patterns are ones every buyer will recognize.

Fig. 3Weeks of supply, slowest first
Fig. 3: Six fictional SKUs with units on hand, units sold per week, weeks of supply and days on hand, sorted slowest first. Lead time is 1 week; the target is 4 weeks.
SKU (fictional)RungOn handSold / wkWeeks of supplyDays
Demo Day Reserve 3.5gSomebody liked the sample a lot.Premium30215.0 wks105
Copycat Cultivars 3.5gSame count as its twin, a completely different problem.Mid40313.3 wks93
Fancy Pants Reserve 3.5gOnly six units, still three months of runway.Top shelf60.512.0 wks84
Boring But Beautiful 3.5gBoring. Boring is beautiful.Mid40104.0 wks28
The Big Bag 28gThe new rung is working. Don’t let it go dark.Value2082.5 wks18
Always Gone Gardens 7gPanic, politely. Reorder now.Value18240.8 wks5

Dashed lines: lead time 1 wk, target 4 wksUnder lead time: reorderNear target: leave itOver 2x target: parked money

A few things jump out once the numbers are in time instead of units.

Boring But Beautiful and Copycat Cultivars both have 40 units. On a units report they look like twins. In weeks they’re not even in the same zip code: Boring But Beautiful turns over in a month, while Copycat Cultivars, which has the personality of a photocopy, sits for a season. This is the “two near-identical SKUs” circle from the ladder exercise, now with a price tag attached. If they’re splitting the same shopper, the copy is losing, and your money is sitting in its box.

Copycat Cultivars has the personality of a photocopy.

Fancy Pants Reserve has the smallest pile on the shelf and still has twelve weeks of supply. Small numbers can hide long tails. A top-shelf SKU that sells one unit every two weeks doesn’t need a reorder; it needs a conversation about whether that rung is working or just present.

And Always Gone Gardens, the one with 18 units that looks fine at a glance, is the actual emergency. The name was a warning. Fast movers on the value rung are where empty shelves happen first, and an empty value rung is the exact gap the ladder piece was about. You fill a missing rung, then forget to watch it, and you’ve rebuilt the hole with extra steps.

The Traps

Five ways the number lies to you.

The formula is simple. The bottom half of it, “average units sold per week,” is where the trouble lives. Here are the usual suspects.

  1. The promo hangover

    Say a SKU normally sells 10 a week, then you run it at 30% off and it sells 50. Average those four weeks together and the system thinks it’s a 20-a-week item. You reorder for a demand that only existed while it was on sale. (30% off is not a strategy. It is also not a forecast.) Pull promo weeks out of the average, or at least look at them separately.

  2. The out-of-stock mirage

    A product sold zero units last week. Did shoppers stop wanting it, or did you stop having it? Zero sales because of zero inventory makes a hot item look dead. Worse, if it’s out of stock right now, the formula tries to divide by almost nothing and hands you a weeks-of-supply number that looks like a phone number. Only count weeks when the product was actually on the shelf.

  3. The first-Saturday fallacy

    A new SKU lands, sells 6 units on its first Saturday, and someone announces it’s a 42-a-week item. It is not. New products need a few weeks of real data before the average means anything. Until then, treat the number as a rumor.

  4. The calendar

    Some weeks are just different. Holiday weekends, the day before Thanksgiving, 4/20, the first of the month. If you calculate your average off a monster week, everything looks understocked. Off a dead week, everything looks overstocked. A longer window (eight to twelve weeks) smooths out the noise; a shorter one (two to four weeks) catches trends faster. Look at both when they disagree, because the disagreement is information.

  5. Forgetting the truck

    Two weeks of supply is relaxed if your vendor delivers in three days. It’s a fire if the next delivery is three weeks out. Weeks of supply never means much on its own. It only becomes “good” or “bad” when you hold it up against how long it takes to get more.

Fig. 4The promo hangover, drawn

Three normal weeks and one sale week. Reorder off the average and you buy for a week that only existed on sale.

30% off is not a strategy. It is also not a forecast.

Order Day

Order the gap, not the gut.

Here’s where weeks of supply stops being a report and starts doing work. Instead of asking “how many should I order?”, ask “how many weeks do I want on the shelf after this order lands?” That target has three ingredients: lead time (how many weeks from placing the order to having it sellable), order cycle (how many weeks until you’ll place your next order for this vendor) and a cushion (a little extra for the busy weekend you didn’t see coming).

Say your vendor delivers in about a week, you order from them every two weeks, and you like one week of cushion. That’s a target of four weeks.

Lead time · 1 wk

order to sellable

Order cycle · 2 wks

until you order from this vendor again

Cushion · 1 wk

the surprise weekend

= 4 weeks

Target: 1 week of lead time, plus a 2-week order cycle, plus 1 week of cushion, equals 4 weeks.

Now the order basically writes itself:

Order = (target weeks × weekly sales) − on hand − already on order

Always Gone Gardens sells 24 a week. Four weeks of that is 96 units. You have 18 on hand and nothing on order, so you order 78. Copycat Cultivars sells 3 a week. Four weeks is 12 units, and you have 40. You order zero, and you now know you’re carrying about 28 units of money that would rather be somewhere else.

Fig. 5Order the gap
Always Gone Gardens24 a week
Order 78
Need 96

24 a week × 4 weeks = 96 needed. 18 on hand (grey), 0 on order.

Copycat Cultivars3 a week
28 over · order 0Need 12

3 a week × 4 weeks = 12 needed. 40 on hand. The hatched part is parked money.

The target won’t be the same on every rung. Fast value SKUs punish you hard for running out, so a little extra cushion there is cheap insurance. Slow top-shelf SKUs punish you slowly, by tying up cash and aging on the shelf, so a leaner target usually makes more sense. The point isn’t to find the one perfect number. It’s to decide your number on purpose instead of letting the last rep who visited decide it for you.

The Takeaway

Read them together.

The ladder

Is something in each spot?

Weeks of supply

Is that something moving?

The ladder and weeks of supply are better as a pair. The ladder asks whether something is in each spot. Weeks of supply asks whether that something is moving. An empty rung is a question. A full rung with fifteen weeks of supply is a question too, just a more expensive one.

None of this takes a math degree. It takes one division problem and the habit of doing it before you trust your gut about a pile of boxes. Sixty units can be nothing to worry about. Twelve units can be an emergency. The number that tells you which is the one measured in weeks.

By John Hirsch

Previously in Buy SheetAnatomy of an Assortment LadderRead it first if you haven’t: this piece is its other half.
ProcureCulture / Buy SheetDays on Hand