The production KPIs that matter in a small food business answer three questions: did we make what we planned, when we planned it; did each batch come out as the recipe says; and what did it cost against what it should have. That's seven numbers: schedule attainment, yield, cost variance, right first time, kitchen utilization, production waste and labor per good unit, all from your batch records and your plan. They're most useful read together, because one problem in the kitchen usually moves several of them. In the example below, a fryer running cool in the second week of October puts five of the seven on watch: three failed batches, yield at 95.1%, costs 4.7% over standard, $214 scrapped and the schedule slipping to 84.6%. Below is each KPI with its formula, how to read them together, how to set targets, and a free dashboard.
The short version
- Schedule attainment: batches made on their planned day ÷ batches planned.
- Yield: good units ÷ units planned; the recipe's promise against what came out.
- Cost variance: actual cost ÷ standard cost − 1; above zero is over standard.
- Right first time: batches with no nonconformance ÷ batches.
- Kitchen utilization: hours used ÷ hours available.
- Production waste and labor per good unit: what scrap and hours cost.
- TaroStack records every batch's yield and cost against standard as it's made. More below.
The seven KPIs
Windward Roots, the made-up taro business I use in these examples, works out its production KPIs at the end of each month. October:
| KPI | Formula | October | Target | Status |
|---|---|---|---|---|
| Schedule attainment | Batches made on their planned day ÷ batches planned | 84.6% | 90% or more | Watch |
| Yield | Good units ÷ units planned | 95.1% | 97% or more | Watch |
| Cost variance | Actual cost ÷ standard cost − 1 | +4.7% | 2% or less | Watch |
| Right first time | Batches with no nonconformance ÷ batches | 88.5% | 95% or more | Watch |
| Kitchen utilization | Hours used ÷ hours available | 80% | 75% or more | OK |
| Production waste | Scrapped at cost ÷ materials used at cost | 4.4% | 3% or less | Watch |
| Labor per good unit | Labor cost ÷ good units | $0.69 | $0.70 or less | OK |
The month's figures: 26 batches planned and 22 made on their day, 3,900 units planned and 3,708 good, a standard cost of $5,980 for those units against an actual $6,260, three batches with a nonconformance, 128 of 160 kitchen hours used, $214 scrapped out of $4,900 of materials, and $2,560 of labor.
Did we make what we planned? Schedule attainment
Schedule attainment counts batches made on the day the plan said. Late batches mean late orders, rushed catch-up batches and, in a food business, product that reaches customers with less shelf life. Twenty-two of 26 is 84.6%. How to build a plan worth attaining is in how to prepare a production plan.
Did it come out right? Yield and right first time
Yield compares good units with what the recipe planned: 3,708 of 3,900 is 95.1%. Right first time counts batches that needed no nonconformance form: 23 of 26 is 88.5%. Low yield on every batch points at the recipe; low yield on a few batches points at what happened on those days. What a nonconformance form records is in nonconforming product form.
What did it cost? Cost variance, waste and labor
Cost variance compares what the good units cost with what the recipe's standard cost says they should: $6,260 against $5,980 is $280, or 4.7%, over. Production waste is what was scrapped, $214 of $4,900 of materials, 4.4%. Labor per good unit, $2,560 over 3,708 units, is $0.69. Yield and cost variance move together: a batch that makes fewer good units spreads the same ingredients over fewer packs. How actual and standard cost differ, line by line, is in batch costing example, and how to find which line moved is in food cost variance formula.
Is the kitchen the limit? Utilization
Kitchen utilization, 128 of 160 hours, is 80%. Low utilization in a rented kitchen is money paid for hours unused; near-full utilization means one bad day can't be caught up. At 80%, October had room, but remaking three failed batches used most of it, which is part of why the schedule slipped. How hours turn into what a kitchen can make is in how to calculate production capacity.
Reading them together: one fryer, five KPIs
In the second week of October the fryer's thermostat started reading high, so the oil ran cooler than the dial said. Three batches of chips came out pale and greasy and failed their check: the three nonconformances, most of the $214 scrapped, and most of the 192 units short of plan. The remaining chips batches yielded less too, which is most of the $280 cost variance. The three failed batches had to be remade, which pushed four batches off their planned days. One thermostat, five KPIs on watch. The fix is a check: record the oil temperature on every chips batch, with limits, so a cool fryer is flagged on the first batch, not found on the third.
Setting your own targets
The targets are Windward Roots' own, not industry figures. Set yours from your last three months, product by product if your products differ much: a recipe that has always yielded 92% isn't failing at 92%, but one that has always yielded 98% is. Work the KPIs out monthly; look at yield and cost variance for each batch as it's recorded, while the cause is still in the room.
Get the dashboard
Download the production KPI dashboard. Type your month's twelve figures on the first sheet, and the dashboard works out the seven KPIs and marks each OK or Watch against targets you set. It's an ordinary Excel file with no macros, and it works in Google Sheets, LibreOffice and Numbers.
Where this stops working
The twelve figures behind the dashboard live on batch sheets: planned and actual dates, planned and good units, the lots used and what they cost, the checks and the scrap. At 26 batches a month, adding them up is a day's work, and by the time October's yield is worked out in November, the thermostat has been wrong for three weeks.
How TaroStack does it
TaroStack records each batch's yield, cost and checks as it's made, so the KPIs come from the batches themselves.
Every batch against its recipe. Production orders carry their due dates, and alerts flag production orders due. Recording a batch takes ingredients off stock from the lots used, at what you paid, records what came out good, and shows its actual cost against the recipe's standard cost. Recipe truth checks, from your own batch history, whether a recipe yields what it claims and which ingredient's quantity is wrong.
Checks that catch the cause. Batch record forms carry your own fields, like oil temperature, with limits, so a reading outside them is flagged when it's entered, and a batch can sit on hold until it's checked. Scrap is recorded with its reason, so waste is reported by reason, item and site.
And the hours, on Standard. Planning on Standard holds work centers and their hours, so a week's plan doesn't promise more than the kitchen has, and suggests what to make and when.
| The KPI | Batch sheets | In TaroStack |
|---|---|---|
| Yield | Added up at month end | Every batch, as it's recorded |
| Cost variance | Worked out in November | Actual against standard, every batch |
| Which ingredient is off | A guess | Recipe truth, from your batches |
| A cool fryer | Found on the third batch | Flagged on the first reading |
| Hours against the plan | A whiteboard | Work centers and capacity, on Standard |
Recipes, batches, actual against standard cost, recipe truth and batch records with limits are on every plan, from $49 a month; planning, with capacity, is on Standard at $99. The first 30 days are free, with no credit card. If you'd like a hand setting up, ask, and we'll do it with you.