Par level calculator for a coffee shop

How do you calculate a par level for a cafe?

A par level is not a target and it is not how much you keep. It is a trigger. It answers one question at the count: is there enough on this shelf to reach the next delivery, and if not, how much do I order. Everything below is that question written as arithmetic.

base       = daily usage x days to cover
safety     = base x safety percent
par        = round up (base + safety)
order to   = round up (par + daily usage x days between orders)
suggestion = order to - what you counted

The same sum with the seeded numbers in it, so you can check the calculator against your own arithmetic:

whole milk, 14 bottles a day, delivery lands the next day

base     = 14 x 1              = 14 bottles
safety   = 14 x 20 percent     = 2.8 bottles
par      = round up 16.8       = 17 bottles

order to = 17 + 14 x 2 days    = 45 bottles

counted 12 on Wednesday: 12 is under par, so order
suggestion = 45 - 12           = 33 bottles

Days to cover means days until the delivery is actually on your shelf, not the number the supplier quotes. If you order Saturday and nobody delivers Sunday, Saturday's cover is two days, not one. That single mistake is the most common reason a par that looks right still runs a shop out of milk.

Safety stock as a percent rather than a fixed quantity is deliberate. A percent scales: it pads a 14 bottle item by about three and a 1,000 gram item by 250, without you writing a separate rule for each one.

Round up, never down, and round to the pack. There is no such thing as ordering 2.8 bottles, and a supplier who sells oat by the case of twelve will not send you seven. Enter your case size and the sums land on numbers you can actually put on an order.

A par is right until usage moves. It is a snapshot of a season, and a shop that sets pars in February and never revisits them is over-ordering hot chocolate in July. Date the sheet.

A worked example on a real shelf

The calculator opens on whole milk, because it is the item most cafes run out of and the one where the arithmetic is easiest to check. Fourteen bottles a day. The truck comes the next morning. Twenty percent padding on top.

That gives a par of 17 and an order-to level of 45. Which means: when the count reads 17 or fewer, the order goes in that day, and it goes in for enough to reach 45. Nobody stands at the reach-in doing mental arithmetic, and nobody has to guess whether 19 is a lot or a little.

Now do it for something slow and far away. Matcha at 100 grams a day, ten days of lead because it ships from a distributor, and 25 percent padding because a stockout there closes a whole section of the menu. That comes out at 1,250 grams. Long lead items get padded harder on purpose: the cost of holding a bit extra is small, and the cost of being out for a week is the drink you sell most of.

Then add both to the sheet. One item is a demo. A cafe stocks about sixty of them, and the sheet is the thing you actually take to the shelf.

Both of those rows, and the rest of the shelf, are laid out in the cafe par sheet template.

Frequently asked questions about par levels

What is a par level in a coffee shop?

It is the count at which you reorder. Not the amount you keep, not a target, and not a maximum. If the par on whole milk is 17 bottles, then the moment a count reads 17 or fewer the order goes in. Above it, you skip. The point is to replace a judgement call at the reach-in with a number anyone on your team can read off a sheet, which is why a par only works if it is written down where the counting happens.

How do you calculate a par level?

Multiply what you use a day by the days until the next delivery actually arrives, add a safety percentage, and round up to a whole pack. Whole milk at 14 a day with a next-day delivery and 20 percent safety is 14 x 1 = 14, plus 2.8, rounded up to a par of 17. Then set an order-to level above it. 17 plus another two days of usage is 45, so you know not just when to order but how much.

What safety stock should a cafe use?

Twenty percent is a reasonable default for something that arrives the next day, and long lead items should be padded harder, 25 to 50 percent, because the downside is not symmetrical. Being a little over on matcha costs you a few days of cash on a shelf. Being out of matcha for the five days it takes to arrive costs you every matcha drink in that window. Pad in proportion to how long you would be stuck, not in proportion to the item's price.

Do lead days mean delivery days?

No, and this is the trap. Lead days are days until it is on your shelf, which includes every day the truck does not run. A supplier who delivers Monday to Saturday has a Sunday hole in it: an order placed Saturday does not arrive Sunday, so Saturday's cover is two days. Cutoffs do the same thing. Miss a 10:50 AM cutoff by ten minutes and next-day became the day after. Count the calendar, not the promise.