Ask three people in a plant what a minute of downtime costs and you will get three numbers, usually large and usually impossible to defend. Somebody multiplies line output by selling price, adds everyone’s wages and rounds up. Finance asks where the figure came from, and the conversation ends there. A number that survives that conversation is built differently. It values lost output at contribution margin, not price. It counts only the labor that is actually idled. It charges restart scrap to each stop. And it is honest about whether the lost output could have been made up. This article walks through that method, which is the same one our downtime cost calculator uses.
The formula
The core calculation has two parts, plus the totals that follow from them:
- Cost per minute = (production rate per hour ÷ 60 × contribution margin per unit) + (crew idled × loaded labor rate ÷ 60)
- Cost per stop = cost per minute × stop duration in minutes + restart scrap units × material cost per unit
- Monthly cost = cost per stop × stops per month
- Annual cost = monthly cost × 12
Every input is something a plant already knows or can measure. Each one deserves a short explanation, because each is where a careless estimate goes wrong.
Why contribution margin, not selling price
When a line stops, you don’t make the units you would have made. But you also don’t spend the material, packaging and variable energy that would have gone into them. What you lose is the difference between the price and those variable costs: the contribution margin. That is the amount each unit would have contributed toward fixed costs and profit.
Fixed overhead doesn’t belong in the number either. Rent, depreciation and salaried staff are paid whether the line runs or not. Adding “fully absorbed cost” or selling price double-counts money that was never at risk, and any finance team will spot it.
Labor: count only who is actually idled
The labor term covers people who are paid during the stop and cannot do other productive work. On a staffed packaging line that might be the whole crew. On a line where operators move to another task during a stop, it may be nobody. Use the loaded labor rate, meaning wages plus benefits and payroll taxes, which your HR or finance team can provide.
Maintenance technicians repairing the fault are a different matter. Their time is a repair cost, not idled labor, and it belongs in the list of costs left out below.
Restart scrap belongs to the stop, not the minute
Many processes make bad product when they restart: a purge, a web that has to be rethreaded, the first containers off a filler, molded parts made before the tool is back at temperature. That scrap depends on how many stops there are, not on how long each one lasts. A 40-second jam and a 40-minute bearing failure can produce the same restart scrap.
This is why frequent short stops cost more than their minutes suggest, and why they deserve proper measurement. We covered that in microstops and the hidden factory. Restart scrap is valued at material cost here, because that is the money actually thrown away.
A worked example
All numbers below are hypothetical, chosen to make the arithmetic easy to follow.
| Input | Value |
|---|---|
| Production rate | 3,600 units per hour |
| Contribution margin per unit | $0.85 |
| Crew idled during a stop | 6 people |
| Loaded labor rate | $42 per hour |
| Typical stop duration | 25 minutes |
| Restart scrap per stop | 150 units |
| Material cost per scrapped unit | $0.40 |
| Unplanned stops per month | 12 |
Now the calculation, step by step:
- Lost margin per minute: 3,600 ÷ 60 = 60 units per minute, × $0.85 = $51.00.
- Idled labor per minute: 6 × $42 ÷ 60 = $4.20.
- Cost per minute: $51.00 + $4.20 = $55.20.
- Cost per stop: $55.20 × 25 = $1,380, plus 150 × $0.40 = $60 of restart scrap, = $1,440.
- Monthly cost: $1,440 × 12 = $17,280.
- Annual cost: $17,280 × 12 = $207,360.
For comparison, suppose the same calculation used a selling price of $2.10 per unit instead of margin. The cost per minute would be $130.20, the cost per stop $3,315 and the annual figure $477,360, more than twice as high. The larger number feels more persuasive in a slide deck. It is also the one that will be taken apart in the budget meeting.
Recoverable or lost? Spare capacity changes the answer
The formula assumes that lost output is lost for good. That is true on a sold-out line running around the clock, where every unit not made is a sale that goes elsewhere. It is not true on a line with spare capacity.
Suppose the line in the example runs two shifts, five days a week, and the 300 minutes lost each month (12 stops × 25 minutes) can be made up on a Saturday. The margin isn’t lost; the units are simply made later. The real monthly cost then looks like this:
- Idled labor during the stops: $4.20 × 300 minutes = $1,260.
- Restart scrap: $60 × 12 stops = $720.
- Overtime to recover the output: 6 people × 5 hours × a hypothetical $63 per hour overtime rate = $1,890.
That comes to $3,870 a month, not $17,280. Both numbers are correct for their situation. The mistake is using the sold-out number for a line that has slack, or the recoverable number for a line that doesn’t. State which situation applies, and remember that it can change with the season, the order book or a new product launch.
What the formula leaves out on purpose
Some real costs are deliberately excluded from the core number because they vary from stop to stop and are hard to attribute:
- Expedited freight to meet a ship date after a long stop.
- Overtime, in the sold-out case where it is used to protect a delivery rather than to recover margin.
- Customer penalties and chargebacks for late or short shipments.
- Repair costs: parts, maintenance labor and contractor call-outs.
- Collateral damage, such as a failed bearing that takes a shaft or a gearbox with it.
- Quality risk on restart beyond the scrap you can count, such as marginal product that ships.
Leaving them out makes the core figure a floor, not an estimate. List them separately as “costs not included”, with an example or two from your own records where you have them. A conservative number that holds up is worth more than a large one that doesn’t.
Presenting the number to finance
A downtime cost becomes useful when finance accepts it. A few habits help:
- Use their inputs. Take contribution margin from the cost accounting system and the loaded labor rate from finance or HR. Then the debate is about the method, not the data.
- Use measured stops. Take stop counts and durations from machine data, not from memory or the shift log, which undercounts short stops.
- State the recoverability assumption in one sentence at the top.
- Show a range. Run the calculation with low, likely and high values for stops per month and duration.
- Don’t annualize a bad month. Base stop frequency on six to twelve months of history where you have it.
- Send the inputs ahead of the meeting. Agreeing on the numbers before the meeting is quicker than arguing about them during it.
Above all, tie the figure to a decision. In the example, a stop costs $1,440. A monitoring program or a spare part that costs a hypothetical $3,000 a month needs to prevent just over two stops a month to break even. That is a question an operations team can actually answer.
From a number to a plan
The cost per stop is most useful when it is set against the stops you could prevent. Unplanned stops on rotating equipment are often preceded by weeks of measurable change in vibration, temperature or current. That is what predictive maintenance and condition monitoring is designed to catch. Frequent short stops are a measurement problem first, and equipment performance and OEE tracking makes them countable.
Before you decide which assets to monitor, rank them by cost per stop multiplied by how often they stop. The short list is usually obvious once the numbers sit side by side. Put your own figures into the downtime cost calculator, write down the assumptions, and you’ll have a number you can defend.