Savings that survive a CFO’s audit

Savings numbers die in predictable ways. The commonest: measuring against the wrong baseline — list price instead of the net price actually being paid, so the “savings” were mostly discounts that already existed. Next: conflating identified opportunity with realized results, so a pipeline estimate gets reported as an outcome and the gap surfaces a year later as a credibility problem. Third: volume effects wearing a price costume — spend fell because admissions fell, and the initiative took the credit.

Behind all three is the same root cause: the measurement was designed after the negotiation, when every incentive favored the flattering number. By then the honest baseline is gone.

The alternative is almost embarrassingly simple. Fix the baseline with finance before the negotiation starts — actual net unit economics, actual volumes, written down. Then, at contract close, capture one number: baseline versus final negotiated terms, with the contract as evidence. One entry, at the moment of signature, while every fact is fresh and the documents are open.

What this replaces is retroactive spreadsheet archaeology: months-later reconstructions from invoices, disputed baselines, and results nobody fully believes including the person presenting them. A single capture at close is auditable because it was recorded as a fact, at the time, against a pre-agreed starting point. That is the entire trick. The discipline is unglamorous, which is why it is rare — and why it is trusted.

Some of the best work in a supply chain year produces no price reduction at all: the vendor proposed an increase and you held it down, or capped an escalator, or killed a surcharge. That is money that would have left the building. It belongs in the record — as cost avoidance, its own savings type, never blended with price reductions.

The separation is what finance is actually checking. A report that says “reductions here, avoidance there, measured against these agreed baselines” survives any review, because every number has a document behind it. A report with one triumphant blended figure invites the question that unravels it. Honest labels are not modesty; they are what lets the good number be believed.