Some decisions show their saving at once and their cost later, in a different place. This page is about telling which ones, before the decision is made.
Nobody is hiding anything. Financial statements record transactions, and they do it faithfully. A sale that never happens is not a transaction, so no statement can show it. The gap is between what is recorded and what is caused.
That gap is why a decision can be reported accurately, look right, and still leave the business with less than it had.
It looks like a subtraction. It works like a multiplication.Six cases. In each, the saving is one line that comes off the total. What it was feeding is where the sum changes.
In each one, the thing being cut was feeding more than one other thing. With one thing relying on it, a 10% cut is a 10% loss. With three, each passes on what it received:
0.9 x 0.9 x 0.9 = 0.729, so a 10% cut is a 27% loss
The books show one line either way, which is why it reads as a subtraction.
Not every cut works this way. In these, what was removed was feeding little or nothing, so the saving holds:
Every figure here is worked out by the calculators from their own example inputs, which are illustrative and not data about any real company. Open any case to see the inputs and change them.
Reducing headcountWhat the research found. A study of about 200 companies found that people leaving by choice rose after a reduction, in some cases by more than the number let go. The effect was smaller where people felt the process was fair and had reasons to stay.
A review of the published studies found the effect of reductions on financial results to be mixed. Some firms improved. Many did not.
When it is the right call. When the work itself has gone, such as a closed product or a lost contract. When capacity is truly idle. When the business cannot carry its costs any other way. In those cases the saving is real, and the calculator will say so.
Requiring a return to the officeWhat the research found. In a six-month randomised trial of 1,612 employees at one technology company, working from home two days a week cut the rate of people quitting by a third. It made no measurable difference to performance reviews or promotions.
A working paper on S&P 500 firms that announced office requirements found no significant change in profitability or market value afterwards, and a fall in employee satisfaction ratings.
The limits of that evidence. The trial was one company and one arrangement. The second study is observational and was circulated as a working paper, so treat it as early evidence. Neither is about working fully from home, where findings are more mixed. And some work can only be done in person.
When it is the right call. When the work needs people together. When newer staff are not learning what they need. When a team's output has measurably fallen and nothing else explains it.
| Decision | Shows at once | Shows later |
|---|---|---|
| Putting off maintenance | The maintenance budget | Downtime, and a chance of one large repair |
| Cutting training | The training budget | Slower work, more errors, people who leave to learn elsewhere |
| Cutting research and development | This year's spend | The product that would have been sold in two or three years |
| Switching to a cheaper supplier | A lower unit cost | Faults, returns, and customers who do not reorder |
| Outsourcing a team | A lower monthly bill | Hand-over cost, slower answers, know-how that has left the building |
| Cutting marketing | The marketing budget | Fewer new customers once the current pipeline runs out |
| Holding less stock | Cash freed from the warehouse | Sales lost when an item is not on the shelf |
Each of these is worked through in the calculator, with figures you can replace with your own. Several come out ahead. Holding less stock and cancelling an unused subscription usually do.
Five questions before any of themIf the answers are "nothing", "most of it", "never", "very little" and "no", the saving is real. Take it.
The example figures in the calculators are illustrative and are not drawn from these sources.
