TL;DR. A lost minute becomes money in three steps, and each step has a different level of proof. Observed money is what happened. Estimated money is what it probably cost. Confirmed money is what finance signed off. Only finance-confirmed money counts for ROI. Keep the tiers separate and the argument moves from the number to the fix.
Why does “downtime cost” start arguments?
Ask three people what an hour of downtime costs and you will hear three numbers. Operations quotes lost output at selling price. Finance quotes margin. Maintenance quotes the repair bill.
None of them is wrong. They are answering different questions with no shared label. The meeting then spends its time on whose number is right, and the fix waits.
The remedy is not one agreed figure. It is an agreed ladder, where each rung says how much you can rely on it.
What is observed money?
Observed money is the fact layer. How many minutes did the line not run, and how many units were not made, in a named period.
It needs no finance input and no opinion. It is a count. If two people measure it from the same records they should get the same answer.
Operations can use observed numbers immediately. They are the base for everything above them, but they are not yet dollars. They are minutes and units.
What is estimated money?
Estimated money applies a rate to the observed fact. Lost units multiplied by contribution per unit gives a range. A tariff applied to wasted energy gives another.
It is useful and it is honest if you label it. Illustrative example: a filler loses 1,400 minutes in a quarter. At an assumed $30 per minute of contribution, that is an estimated $42,000. Change the rate to $25 or $35 and the answer moves to $35,000 or $49,000.
That range is the point. Estimated money has bounds. It can include a run-rate, but it should never be stretched beyond the period it was observed.
Operations and planning can use it to rank losses and size a case. It is not yet a number for the board.
What is confirmed money?
Confirmed money is a saving that finance has accepted. The finance controller reviews the evidence and signs it off.
The unit of value is one finance-countersigned, verified saving attributed to a named action. Not a forecast, not a hope, not a run-rate multiplied by twelve.
This is the only tier that feeds ROI. If a board pack quotes return on a programme, the numerator must be confirmed money. Anything else is a projection and should say so.
How does a saving move up the ladder?
A saving passes through four evidence states.
- Observed. The loss is a counted fact.
- Adjusted. Known distortions, such as a planned shutdown or a different product mix, are accounted for.
- Supported. The improvement is checked against similar production, not just a calendar month.
- Finance accepted. The controller signs it.
Each step costs effort. That is deliberate. A saving that has not climbed the ladder stays in the tier it has reached, with its label.
What goes wrong when the tiers are mixed?
The common failure is a savings report that adds observed minutes, estimated dollars and hoped-for run-rate into one headline. It looks impressive until finance asks how it was derived.
Then the whole report loses credibility, including the parts that were solid. One unlabelled estimate can cost the programme its next budget.
A second failure is the reverse. Finance refuses everything that is not signed, so operations stops reporting progress. The tiers fix both: operations reports observed and estimated, finance owns confirmed, and everyone sees the label.
How should operations and finance work together?
Agree the rates in advance. Finance owns the contribution per unit and the tariffs. Operations owns the counts. If the rate changes, the estimate changes, and nobody is surprised.
Agree the acceptance rule in advance too. What evidence will finance need to accept a saving? A named action, a before and after on comparable production, and enough time to see whether it held.
What does each tier look like in a real report?
Take one line in one quarter. The observed line reads: “Filler 2, 1,400 minutes lost, 63,000 units not made.” No dollars appear.
The estimated line reads: “Estimated cost $35,000 to $49,000, at $25 to $35 per minute, rate supplied by finance.” The range and the source of the rate are on the page.
The confirmed line reads: “Finance confirmed $31,000, against the capping station repair, verified on comparable runs.” Every figure in this paragraph is illustrative.
A reader can see at a glance which line to trust for which purpose. That is all the ladder does.
What do you say when someone asks for one number?
Executives often ask for a single figure. Give them the confirmed number first, followed by the estimate for what is still open, and name the tier each time.
Resist the temptation to blend. A blended number cannot be defended, because nobody can say how much of it is fact. A pair of labelled numbers can be defended line by line.
If the confirmed number is small, say so. A modest, signed saving builds more credibility than a large, unsigned one, and it grows each quarter as more evidence arrives.
How often should the tiers be reviewed?
Observed money updates continuously, because it is a count. Estimated money should be reviewed whenever a rate changes, which for most plants means once or twice a year, or after a major product change.
Confirmed money is reviewed when a saving reaches enough comparable production for finance to sign. That may be monthly or quarterly depending on how often the loss occurs.
Put these reviews in the finance calendar. Otherwise the ladder is a good idea that nobody has time to climb.
What about savings that finance rejects?
Finance will sometimes decline. The evidence may be thin, the comparison may not be like for like, or a planned event may explain the change.
Treat that as information, not defeat. The saving stays in the estimated tier with a note on what would change the answer, and the owner gathers it.
Over time, teams learn what finance needs, and submissions get stronger. The ladder works because the top rung is hard to reach.
Why does the label matter more than the precision?
A figure to the nearest pound looks authoritative and may be wrong by a third. A rounded range with the right label is more honest and easier to defend.
Readers forgive an estimate that says it is an estimate. They do not forgive a precise-looking figure that turns out to be one. Over a year, that difference decides whether finance trusts operations’ numbers or checks each one.
The label costs a few words. The lack of one can cost a programme.
What to do on Monday
- List every downtime figure in your last report and write its tier next to it.
- Agree with finance the rate used to turn minutes into dollars.
- Separate any “annualised” figure and label it an estimate.
- Ask finance what evidence it would need to confirm your largest claimed saving.
- Quote ROI only from confirmed money.