EOFY is closing in. Most procurement teams are heads-down on year-end. Renewals that were always going to land in May and June are landing now. The phones are about energy contracts, insurance, telco, freight. The team is busy. The CFO is busy. The number that will sit on the FY26 close pack is being assembled.

In that environment, this piece is the case for one quiet exercise that very few mid-market finance teams run, and that almost always pays for itself before the new FY begins. The savings persistence audit.

Here is the pattern I keep seeing in the engagements I have worked on. A mid-market business runs a procurement exercise in some prior cycle. Suppliers are tendered. Rates come down. A savings number is calculated, validated, signed off, and reported. The board sees it. The next quarter the topic moves on. The CFO files it under done.

A year and a half later, the same cost lines are back where they started. Sometimes higher. And nobody noticed, because nothing was watching.

The reason is structural, not behavioural. A procurement exercise is a savings event. It happens once. It produces a number. Then it is over. But cost does not stay down on its own. Volumes drift up as the business grows or activity recovers. Maverick spend creeps back as individual managers reach for their preferred suppliers. Contract terms that were tight at signing get quietly eroded by variations and scope changes that no one tracks against the original baseline. The supplier who sharpened the pencil to win the tender lets the rate float back toward where it was, one quarterly index adjustment at a time. None of it is dramatic. That is exactly why it works.

What separates the businesses whose savings stick from the ones who watch them leak is not a sharper negotiation. It is that someone built a savings engine. The unglamorous machinery that holds the line after the event is over. In practice that machinery is three controls, and most mid-market finance functions run none of them.

The first is contract compliance monitoring. Someone, monthly, checking that what was agreed is what is actually being charged. Not a quarterly glance from a category manager. A monthly read against the signed schedule. The second is demand discipline. Quiet controls on who can buy what, so the savings are not undone by volume creeping back up as the business gets comfortable again. The third is consolidation governance. Making sure the supplier base that was deliberately rationalised does not slowly re-fragment as departments quietly add their own preferred vendors back in.

None of those three are exciting. All of them are where the money actually stays.

The savings persistence audit is the simplest possible test that the engine exists. Pull the savings number reported in the last completed procurement cycle. Pick the categories that made up the bulk of it. For each one, compare last cycle's signed rates and volumes against the rates and volumes the business is paying today. Not in a tender, not in a board pack, in the actual invoices in May. The gap between what was reported and what is being paid is the leak.

Two things tend to come out of that exercise, run quietly in the weeks before EOFY. The first is a real number for the FY26 plan. Not a hopeful figure carried forward from last cycle, but the cost base the business is actually carrying into the new FY. That alone is worth the effort, because the FY26 budget is otherwise built on a savings assumption that is no longer true. The second is the start of a standing rhythm. A monthly check, against last cycle's signed rates, by a named person, with a column on the management report. Not a project. Not a transformation. A standing rhythm that quietly stops the next savings event from drifting back into the P&L the way the last one did.

The harder version of the question is this. If a board director asked you tomorrow which of your supplier savings from the last cycle have actually proven to be still in the P&L today, would the answer be a confident yes, or a pause and a request to come back? If the honest answer is the second, you are in very good company. And the weeks before EOFY are the cheapest, lowest-disruption cost work available to you this year. No supplier is renegotiated. No headcount is touched. No project is launched. Someone simply looks at the invoices against the contracts, and tells the truth about what they show.

The question worth taking into your next leadership meeting before 30 June: which of our supplier savings from the last cycle have we actually proven are still in the P&L today? If the honest answer is uncomfortable, the work that follows is the most accessible cost work available to you in the new FY.

A mid-market manufacturer in Melbourne squeezed its suppliers hard last year and reported the savings to the board. This month, with the year just closed, no one could tell me how much of it was still in the P&L.

That is not poor management. It is how the system is built. A procurement exercise is an event. It has an end date, a number to report, a sense of being done. There is no equivalent ceremony for the slow erosion that follows.

And it does erode. Volumes drift up. Preferred vendors creep back in, one manager at a time. Tight contract terms get softened by variations no one tracks. The supplier who sharpened the pencil to win the work lets the rate float back toward where it started. None of it is dramatic. It is only obvious in the rear-view mirror.

The businesses whose savings stick are not better negotiators. They run three unglamorous disciplines, monthly. Was what we agreed actually what is being charged. Did volume quietly creep back up. Did departments re-add the suppliers we deliberately cut. Most mid-market finance functions run none of the three.

So a question for the new financial year. Not what did we save last cycle. Which of those savings have we proven are still in the P&L today.

A procurement exercise is a savings event. Keeping the savings is a savings engine. Most businesses only ever build the first. Which of last cycle's savings have you actually proven are still there.