Ask a chief financial officer how their last cost program went and the answer is usually some version of "well, at first." The targets were announced, the first-year savings were booked, the program office was thanked and disbanded. Then, somewhere in the second year, costs started to creep back. By year three, a new program was being discussed.
We have reviewed the results of more than sixty cost programs across our client base since 2019, comparing the savings each one reported with what appeared in the accounts two and three years later. On average, programs had kept about 60 percent of their reported savings by the end of year three. The best kept nearly all of them. The worst kept less than a quarter.
The difference between the two groups was not ambition, sector or the size of the target. It came down to a handful of choices made at the start.
Savings were counted, not budgeted
The most common failure is also the simplest. A saving is recorded when a decision is taken, such as a contract renegotiated or a role removed, and reported to the board in a program tracker. Nobody reduces the budget of the department concerned. A year later the money has been spent on something else, and the tracker and the general ledger tell different stories.
Programs that kept their savings did the opposite. Every saving was written into the following year's budget, signed off by finance and owned by the line manager whose budget fell. The tracker was retired as soon as the budget became the record.
The work did not change
Removing a cost without removing the work behind it is a loan, not a saving. If a team of twelve is reduced to ten and the volume of work stays the same, overtime rises, contractors appear, or service slips until headcount is quietly restored.
The durable programs treated cost as a symptom. They asked why the work existed, whether customers would pay for it and how it could be done with fewer steps. In one insurer, a third of the claims team's effort turned out to be correcting errors introduced earlier in the process. Fixing the upstream form saved more than any reduction in the claims team itself would have.
The program office left too early
Program offices are expensive and slightly resented, so there is pressure to wind them down once targets are met. That is usually the moment risk is highest. New budgets are being negotiated, managers who lost headcount are asking for it back, and the executive team's attention has moved on.
The savings that survive are the ones somebody is still watching in month eighteen.
The programs in our top group kept a small team, often two or three people inside the finance function, reviewing cost against the new baseline every month for at least two years after the program formally ended.
What separates the programs that hold
The table below compares the programs that retained more than 85 percent of their savings with those that retained less than 50 percent.
Practice | Held savings | Lost savings |
|---|---|---|
Savings written into next year's budget | 92% | 31% |
Process redesigned, not only headcount reduced | 78% | 24% |
Cost tracking continued for 24 months or more | 71% | 12% |
Incentives tied to cost outcomes below the top team | 64% | 19% |
None of these practices is complicated. All of them require someone to insist on them when it would be easier not to.
Questions for the next program
If a new cost program is on your agenda this year, a few questions asked at the start will tell you most of what you need to know about whether it will last:
Where will each saving be recorded, and who has to agree before it is counted?
For each saving, what work stops, and who confirms it has stopped?
Who will still be tracking cost against the new baseline in two years, and do they have the authority to escalate?
Which managers below the executive team are rewarded for keeping cost down, and how?
None of this makes a cost program more pleasant. It does make it more likely to be the last one for a while, which is the real prize. A company that has to cut costs every three years spends much of its management attention on the same problem, and its employees learn to wait each program out.
