The target is not the purpose
A cost reduction programme is commissioned to protect a business, not to shrink it. That distinction is easy to lose once a target has been set and a workstream has been formed. Teams begin to compete on percentage reductions, the finance function starts counting, and the question of what the business needs to keep doing well quietly disappears from the agenda.
The consequence is familiar: a programme that delivers the number in the year it runs, and a business that is slower, less capable and more fragile two years later. Redundancies in customer service reduce cost per transaction and increase churn. Deferred maintenance reduces spend this year and produces a capital bill next year. Cut training and you buy recruitment.
The test for every saving should be the same: does this reduce cost without reducing the capability the business depends on? Savings that fail that test are not savings, they are borrowing.
Where the money usually is
Start with spend that has no owner. Indirect procurement is the most reliable source of savings in most mid-sized businesses: print, marketing, software licences, facilities, professional fees, telecoms, travel and the long tail of subscriptions that nobody remembers approving. These categories are rarely negotiated, frequently duplicated and almost never benchmarked.
Then look at the shape of the organisation rather than the number of people. Layers of management, spans of control, duplicated roles after an acquisition, and processes that require three approvals for a decision that used to need one. Redesigning the structure can remove cost without removing the people who do the work.
Then the estate. Space utilisation after a change in working patterns is often poor, and a lease event is the moment to fix it. Then systems: overlapping platforms, unused modules, contracts that renew automatically at last year's price.
Direct labour is usually the largest line and the last place to look, because it is the easiest to cut and the hardest to reverse. If the other categories have not been examined first, the programme is not a cost programme; it is a redundancy exercise with a longer name.
The cuts that come back as costs
There is a set of reductions that reliably reappear elsewhere in the accounts. Reducing maintenance increases breakdowns and emergency spend. Reducing quality control increases returns and complaints. Reducing sales capacity reduces revenue next year. Reducing finance headcount increases the cost of external accountants. Each looks like a saving in the month it is taken and behaves like a transfer later.
The second category is capability loss that is invisible until it is needed. The person who knew how the pricing model worked, the team that could deliver the complex project, the supplier relationship that took years to build. When those go, the business does not notice until a customer asks for something unusual.
The discipline is to identify the capabilities the strategy depends on before the cuts are designed, and treat them as protected. Everything else is fair game. Without that step, the programme optimises for the metric and damages the business.
Sequencing, governance and communication
Establish the baseline properly. A savings target calculated from a budget is a target; a baseline calculated from the actual run rate is a plan. Reconcile the two before workstreams begin, or the programme will spend its first quarter arguing about the starting point.
Give the programme one owner with real authority, a small central team, and workstreams aligned to categories of spend rather than to departments. Set decision rights explicitly: who can approve a contract change, who can authorise a restructure, who arbitrates when two workstreams claim the same saving.
Then communicate honestly and early. People accept difficult decisions when the reason is clear and the process is fair; they resist when they hear about it from a supplier. The cost of poor communication is not measured in goodwill, it is measured in the productivity lost while everyone works out what is happening.
Proving it worked
Savings that cannot be seen in the accounts are not savings. Track them in the management accounts, month by month, against the baseline, with the cost centre and the owner identified. Anything not visible after two months should be treated as unrealised.
Watch the counter-measures as closely as the savings: service levels, customer churn, staff turnover, overtime, contractor spend, rework and complaint volumes. If the savings line improves while two of those deteriorate, the programme is moving cost rather than removing it.
Finally, hold the gains. The most common failure of a cost programme is that the savings are delivered and then quietly spent back over the following eighteen months, because the controls that protected them were removed with the programme team. The controls are the deliverable, not the report.