The argument

The final quarter is the worst possible time to redesign an operating model and the best possible time to prove one. The constraint is the point: with weeks rather than months, you cannot restructure your way to a result. You can only remove friction that is already visible.

Almost every Q4 efficiency push we are asked to help with begins with the same instinct — cut something. Cost comes out, the number lands, and by March the capability that was quietly cut has to be rebuilt at a premium. The alternative is not softer. It is more specific.

What follows is where the recoverable time and money actually sit in the last twelve weeks of a corporate year, in the order we look for them.

Before anyone counts people, count waiting. In most corporate processes the work itself takes hours while the process takes weeks, because the item sits between steps waiting for an approval, a handover, or a meeting that only happens on Thursdays.

Map one high-volume process end to end and mark, honestly, how much of the elapsed time is work and how much is waiting. A ratio of one to eight is entirely normal. The waiting is free to remove. The work is not.

It also changes the conversation. A team told to be twenty per cent more efficient will protect its people and cut its discretionary spend. A team shown that its own process spends eleven of thirteen days waiting will usually redesign the process itself, and do it better than an outsider could.

An approval threshold set in a different era is one of the most expensive artefacts a company carries. A sign-off limit that made sense a decade ago now routes several thousand transactions a year through a director who approves ninety-eight per cent of them without discussion.

Raise the threshold, sample the exceptions afterwards, and you convert a control that was never controlling anything into days of recovered management time — without touching a single role.

In practice

Ask an operations team to list the recurring reports it produces, then ask who read the last one. The honest answer usually retires a third of them inside a week, along with all of the collection work sitting behind them.

The rule we use is unsentimental: a recurring report survives only if a named person can describe a decision they made differently because of it. Everything else becomes available on demand instead.

Cost comes out fastest where nobody is looking — the waiting between steps, the approval that approves everything, and the report that nobody reads.

The best fourth-quarter work is often not about the fourth quarter at all. December is when the year’s operational debt is at its most visible and least contested: the workaround everyone tolerates, the spreadsheet holding a critical process together, the supplier nobody has had time to renegotiate.

Fixing one of those in the quiet fortnight at the end of the year buys back the whole of the following one. It will not show up in this year’s number, which is precisely why it never gets done.

Efficiency programmes fail on ordering far more often than on ambition. Removing a step before you understand why it exists creates a compensating step somewhere less visible, and the saving reappears as an exception queue two functions away.

Take one process fully to the end before starting a second. A single closed loop — measured, changed, measured again — gives you a credible number and a template. Six half-finished ones give you a slide.

The cost that comes out easily in December is usually the cost that was doing quiet work: the second supplier kept warm, the training budget, the analyst whose whole job was noticing things early.

Before any reduction is signed, ask what it was insuring against and what will happen the first time that risk shows up. If nobody can answer, the line was genuinely redundant. If everybody can, you are not saving money — you are borrowing it at a rate you have not been quoted.

Treat the last quarter as a diagnostic rather than a deadline. Remove waiting before headcount, retire the controls that no longer control anything, and spend the quiet weeks on the debt that would otherwise define your January. The number lands either way. Only one of those routes leaves the organisation stronger in February.