A Printable Par Sheet Template for Your Cafe
Mark, founder of Parly·September 2, 2026·6 min read
The column everyone leaves blank
Every count sheet template you download has a par column. Almost every one you find in a working cafe has that column empty, or filled once, in pen, eighteen months ago, by someone guessing.
That is not laziness. It is that the templates hand you the column and never tell you what number goes in it. So the column sits there looking official while the actual ordering decision happens the way it always did: you look at the shelf, you feel like that is not enough oat milk, you order two cases.
A par is not a feeling about a shelf. It is a specific number with a specific job: the level at which you must order this item today or you will run out before the next delivery lands. Below par means order. Above par means wait. That is the whole idea, and once the column is filled correctly the shelf stops needing a judgment call.
Here is the sheet, and here is the arithmetic that fills it.
What a par sheet has to hold
A par sheet is not a count sheet with an extra column. It answers a different question, so it holds different things. Six columns, and you need all six.
| Column | What it holds | Why it earns the space |
|---|---|---|
| Item | The thing as you count it | "Oat milk", not "alt milks" |
| Count unit | What one of it means | A bottle and a case are not the same 1 |
| Daily usage | How much leaves per day | The engine of the whole sheet |
| Lead days | Order to delivery, in days | Comes from your supplier, not your habit |
| Par | Order at or below this | Usage times lead days, plus safety |
| Order to | Where you bring it back up to | Par plus a full order cycle |
Two of these are worth arguing about.
Count unit is where most sheets quietly break. If your count says "oat milk: 3" and your order says "oat milk: 3," one of those is bottles and the other is cases of twelve, and you have just ordered four times what you needed. Write the unit next to the item and never carry two units for one item on one sheet.
Order to is the column that separates a par sheet from an alarm. Par tells you to order. It does not tell you how much. Without an order-to level you hit par, panic, and buy a number that felt right, which is exactly how over-ordering happens even in shops that count religiously.
The math that fills the par column
Three inputs, one line of arithmetic.
Daily usage comes from your own numbers, not a benchmark. Two ways to get it. If the item maps to recipes, run sales times recipe: my milk drinks tell me whole milk runs about 14 bottles a day and oat about 13.5. If it does not map to a drink, which covers most of your paper and chemicals, use count depletion: last Monday you had 12, Wednesday you had 6, so six left the shelf over two days, so three a day.
Lead days is the honest gap between placing the order and the box landing, and it is a supplier fact you look up once. My dairy cutoff is 4:50 PM for next-day delivery, so lead is one day. My paper supplier cuts off at 10:50 AM, also next day. Matcha is five to seven business days, which is nine or ten calendar days, an entirely different animal that deserves its own ordering habit.
Safety is a percentage on top, and it is where judgment belongs. I run about 20 percent on things that move fast and predictably, more on anything with a long lead or a supplier who has missed a delivery on me. Safety is not padding for a bad estimate. It covers the Saturday that runs 30 percent over a normal Saturday, and nothing else.
Worked, on the two items you will use it on most:
Whole milk. 14 bottles a day. Lead 1 day. Par = 14 x 1 = 14, plus 20 percent safety, so 17. When the Wednesday count reads 17 or fewer bottles, the order goes in before 4:50 PM. That is it. No standing at the reach-in.
Matcha. Say your count history shows 100 grams a day. Lead is seven business days, call it ten calendar days to be honest about weekends. Par = 100 x 10 = 1,000 grams, plus 25 percent for a long-lead item, so 1,250. Notice what that does: a fast item pars out at one day of stock, a slow-shipping item pars out at nearly two weeks. Same formula, wildly different answers, and this is precisely why one blanket "keep two weeks of everything" rule wastes cash on milk and still runs you out of matcha.
⚠ Lead days are not delivery days
An item from a supplier who only delivers Monday through Saturday needs the Sunday gap inside its lead number. Count the days the truck does not run, or your Saturday par is wrong every weekend.
The printable par sheet
Print this, or paste it into a spreadsheet. The items are the ones a specialty cafe actually stocks, in walk order so it matches the sheet you count from. The usage and par numbers are examples from a shop about my size; overwrite every one of them with yours.
| Item | Unit | Usage/day | Lead days | Par | Order to |
|---|---|---|---|---|---|
| Whole milk | bottle | 14 | 1 | 17 | 45 |
| Oat milk | case of 12 | 1.1 | 1 | 2 | 6 |
| Half and half | bottle | 1 | 1 | 2 | 6 |
| Heavy cream | bottle | 1.5 | 1 | 2 | 8 |
| Ethiopian beans | 5 lb bag | 0.8 | 5 | 5 | 15 |
| Peru beans | 5 lb bag | 0.6 | 5 | 4 | 12 |
| Matcha | gram | 100 | 10 | 1250 | 3000 |
| Chai concentrate | bottle | 1 | 7 | 9 | 20 |
| 16 oz cups | sleeve | 2 | 1 | 3 | 12 |
| Hot lids | sleeve | 2 | 1 | 3 | 12 |
| Cafiza | jar | 0.05 | 1 | 1 | 3 |
| Nitrile gloves | box | 0.3 | 1 | 1 | 4 |
Two things to notice before you copy it. The par on a case item is a small whole number, because one case is a lot of days; that is normal and it means the item only ever needs checking once a week. And the chemicals sit at a par of 1 with an order-to of 3, which is correct: a jar of cafiza lasts a month, and there is no reason to carry six.
Pars move, so put a date on them
A par is true for the season you set it in. Iced season moves cups, oat, and matcha; the first cold week moves whole milk and hot lids the other way. A par set in February and never touched is wrong by June in both directions at once, which is how a shop ends up with too much of one thing and short of another while counting perfectly the whole time.
Put a review date in the sheet header and redo the math quarterly, and again whenever you change the menu or a supplier changes a cutoff. It takes twenty minutes: pull the last four weeks of usage, rerun usage times lead days, and overwrite the column. If you want the reasoning behind the reorder point itself rather than the sheet, par levels are explained in full here.
One habit makes the whole thing hold: read the par column at the count, not at the order. The number is only useful in the moment you are looking at the shelf. Written into the count and read out loud, it turns the order into arithmetic instead of a memory exercise.
Fill in two rows tonight
Do not fill the whole sheet. You will not finish and the file will die in Downloads.
Pick your two fastest-moving items, almost certainly your default milk and your espresso bean. Pull the last two counts and work out what left per day. Look up the supplier cutoff and lead time from the last order confirmation email. Multiply, add a fifth, and write the number in the par column.
Then at Wednesday's count, read those two par numbers before you decide anything. If the count is at or below par, order to the order-to level. Do that for two weeks on two items and you will fill the other thirty rows without being asked, because the column will have started making the decision for you.