Bakery Production Planning: How to Decide How Much to Bake Every Day

· 7 min read

A baker kneading dough by hand on a floured counter

TL;DR

  • The method: plan each product for each day from the same weekday's recent sales, corrected for sell-outs and leftovers, then adjusted for known events.
  • Why averages fail: in our example, baking the weekly average every day leaves 27 croissants over on Monday and 57 short on Saturday.
  • The signal most bakeries lose: when something sells out, your sales figure understates demand. Record the sell-out time.
  • Paper works: under a dozen products and one shop, a weekly sheet is enough. Software earns its place when the product list or the number of shops grows.

How do you decide how much to bake each day?

Plan every product separately, for each day of the week, starting from what that product sold on the same weekday recently. Then correct that number for the days it sold out (demand was higher than sales) and the days it was left over (demand was lower), and adjust for anything you know is coming: a holiday, a market, a forecast heatwave.

Five steps, and none of them needs software. Here they are with one product, the butter croissant, as the example. The numbers are illustrative.

Step 1: Start from the same weekday, not the weekly average

Here are four weeks of croissant sales, averaged by weekday:

Mon Tue Wed Thu Fri Sat Sun
Croissants sold (4-week average) 92 88 95 101 128 176 150

The average across the week is 119. A bakery that bakes 119 every day ends Monday with 27 left over and runs out on Saturday 57 short. Both cost money: the first in ingredients and labor, the second in sales that walked out the door.

Use the median of the last four same weekdays rather than the average if one of them was unusual. A public holiday Monday should not drag every following Monday's plan up with it.

Step 2: Correct for the days it sold out

A sales figure only tells you what you sold, not what customers wanted. If croissants were gone by 1:15 pm on Saturday, 176 is what you had, not what people would have bought.

Record the time each product sells out. It is the single most useful number a bakery can add to its daily routine, because it turns "we sold everything" into "we sold everything by 1:15, with four hours still to go."

A practical rule:

  • Sold out in the last hour or two before close: the quantity is about right. Keep it.
  • Sold out well before that, several weeks in a row: raise that day's quantity by a step, around 10%, and check the sell-out time again next week.

Avoid estimating the missing demand by scaling up to the full day. Sales are not spread evenly across opening hours: a morning rush and a quiet afternoon break any straight-line guess. Move in steps and let the sell-out time tell you when to stop.

If your till records sales by time, you can do better than steps. Work out what share of a normal day's sales your shop has usually made by each time of day. If croissants sold out at 1:15 pm and a normal Saturday has made 70% of its sales by then, 176 sold suggests demand of about 176 ÷ 0.70 ≈ 250. Treat that as an upper bound rather than a target, because an early sell-out extrapolates from a small part of the day.

Step 3: Correct for the days it was left over

The reverse applies. If Monday regularly ends with around 11 croissants unsold and they never sell out, bring Monday's quantity down by roughly that amount, then watch whether they start selling out early.

Count leftovers per product, at close, every day. A single "we threw out about a bin" at the end of the week cannot tell you which product or which day to change.

Step 4: Adjust for what you know is coming

Your recent history cannot know about next week's events. You can:

  • Holidays and local events. Look at the same event last year for each product rather than applying one percentage to everything. A long weekend can double sandwich loaves and halve office-lunch pastries.
  • Weather. Its effect is real but product-specific: rain can lift hot drinks and comfort pastries while emptying a terrace. Only adjust for weather where your own history shows a pattern.
  • Orders. Take pre-orders and catering out of the walk-in plan and add them on top, so a one-off order of 80 croissants does not inflate next month's plan.

Step 5: Turn it into a production sheet

The output is a sheet per day: every product, the quantity to make, and ideally which bake it goes in. Splitting high-volume items into two bakes (a morning bake and a smaller late-morning one decided by how the morning went) is one of the most effective ways to cut leftovers without selling out early.

What production planning software adds

The method above works on paper. What it costs is time: someone has to pull each product's sales for the same weekday, remember last week's sell-out times and leftovers, and write the sheet, every day.

That is the job we built Baikely to do, and it runs the same method. For each product and each day, it starts from the same weekday over the last three weeks. On days the product sold out, it adds an estimate of the sales you missed, based on your own shop's sales rhythm through the day and capped so an early sell-out cannot inflate the plan. Special days are planned from the same day last year. Next to each suggestion it shows last week's sales, what sold out and when, what was left over and the weather, so you make the final call in a few seconds:

The daily plan: last week's sales, sell-out times and leftovers next to each suggestion. "Accept all" confirms the day in one click.

The same plan is laid out for the whole week, so you can adjust any product on any day and see the totals move:

The weekly plan, with the forecast for each day. Edited quantities turn green once saved.

Do you need software or is a spreadsheet enough?

Honestly, a spreadsheet is enough when:

  • you sell fewer than about a dozen products,
  • from one shop,
  • and the person who plans also bakes, so sell-out times and leftovers are remembered, not recorded.

Software starts paying for itself when the product list grows, when a second shop or a second baker gets involved, or when nobody has the half hour a day the sheet takes. The test is simple: if your leftovers and your early sell-outs are not both going down month after month, the plan is not getting the feedback it needs.

Frequently asked questions

How far ahead should a bakery plan production?

Plan the week ahead so ordering and staffing line up, then confirm each day's quantities the day before, when you know the latest sales, the weather forecast and any new orders.

What is a par level in a bakery?

A par level is the standard quantity you make of a product on a given day. It is where most bakeries start. The method above is a way to keep par levels up to date per weekday instead of setting them once and forgetting them.

How much leftover is normal?

Some is healthy: a product that never has any leftover is probably selling out and losing sales. The target is small leftovers and sell-outs close to closing time, not zero. Our guide to bakery waste covers how to measure both sides.

Does production planning change my prices?

Indirectly, yes. Every unsold item raises the cost of the ones you did sell. See how to price baked goods for the worked example.