Problem, then fix

How much flour to order, and when

Running out of flour on a Saturday is rarely a supply problem. The mill delivered on time. The sacks were counted last week. What went missing was a small piece of arithmetic that nobody had time for on Thursday, because the numbers it needs live in three places.

What the arithmetic needs

Three things, and every bakery already has them.

How much flour each product carries. A 900 g loaf at 75 percent hydration holds about 450 g of flour; a 90 g roll about 50 g. The recipe’s percentages give this number for every product.

What the coming week will bake. The bake list: 480 large loaves, 360 small, 900 rolls, 60 trays of focaccia. Multiply each line by its flour, add them up, and the week’s flour use falls out. Add a few percent for dusting, starter feeds and waste.

What is on the shelf, and when the mill delivers. A count of the store room, the delivery days, and the cut-off for orders.

The two dates that matter

Put those together and two dates appear.

The run-out day: stepping forward from today’s stock, subtracting each day’s bake, the day the flour reaches zero. Fridays and Saturdays take a bigger bite than Tuesdays, so the step is not even, and that is exactly why “about a week” is not good enough.

The order-by day: the run-out day minus the mill’s lead time, minus a margin of one delivery gap so that a late lorry costs a phone call rather than a Saturday.

When the order-by day is today, you find out on Monday morning. Without the arithmetic, you find out on Saturday, from the person reaching for the sack.

Why this goes wrong in real bakeries

Each input drifts on its own. A new wholesale account adds sixty loaves on Thursdays and nobody redoes the sum. Sacks get used from a second location, or a delivery arrives short, and the count is stale. The mill moves its cut-off for a bank holiday.

Any one of these is easy to handle. All three, weekly, by hand, is the reason the calculation does not get done, and the reason the pallet gets checked on Friday instead.

The flour that surprises you

It is seldom the main white flour. It is the strong flour that goes into one product, or the seed mix, or the rye: a small weekly figure, a small stock, and no one watching it. Any forecast that only covers the main sack misses the ingredient that actually runs out.

Doing it without doing it

The inputs above already exist in your recipes, your bake list and your supplier’s calendar. Joined together, they produce the run-out day and the order-by day for every ingredient, every morning, without anyone adding anything up. The free flour forecast does this with numbers you type in for your five or six main ingredients. bakeryOS does it from the production plan itself and drafts the supplier order on the day it is due.

Questions bakers ask

How much flour does a small bakery use per week?

A microbakery baking 150 to 250 loaves a day uses roughly 400 to 800 kg a week. A bakery with a wholesale round can use two to five tonnes. Your own number follows from the bake list, because every loaf carries a known amount of flour.

How do bakeries know when to order flour?

The good ones work forward from today's stock and the coming week's bake list to the day the sack runs out, then back from that day by the supplier's lead time. Everyone else looks at the pallet on Friday.

What is a par level?

A minimum amount to keep on hand, chosen so a normal delivery cycle never empties the store. It is a rough stand-in for a forecast, and it works while the bake stays the same every week.

Related

• Flour protein content for bread: what the number means and what to buy. Bread flour runs 11.5 to 13.5 percent protein; under 11 is pastry, over 14 suits enriched doughs and long fermentations. Protein quality matters as much as the number.
• Baker's percentages, explained simply. In baker's percentages, flour is always 100 and every other ingredient is a share of the flour. Water 70 means 700 g of water per kilo of flour, at any batch size.
• Bakery production planning: when a spreadsheet stops being enough. A spreadsheet plans a bakery well up to about fifteen products and a few standing orders. Past that, the bake list, mixing sheet and stock start to disagree.

Sources: Gisslen, Professional Baking; Own practice: a six-day sourdough microbakery in Helsinki. Updated 2026-09-11.