Why Finance Never Believes the Benefits Case — and How to Make Programme Numbers Bankable
The CFO's disbelief is not obstruction. It is arithmetic.
The Question That Ends the Meeting
Every programme director learns, sooner or later, to dread a particular kind of silence. It arrives in the benefits review, usually around month fourteen, after the milestone slides have shown green and the steering group has begun to relax. The programme has reported its number — let us say eight million pounds of recurring annual savings, now formally “realised”. And the finance director, who has contributed little to that point, asks a single question: if that is true, why is next year’s cost base identical to this year’s?
There is seldom a good answer in the room. The benefits manager reaches for the tracker, the sponsor reaches for the narrative, and the number that felt solid in the business case begins, visibly, to soften. Everyone present understands what has happened, even if no one says it aloud: finance does not believe the figure, and finance controls the ledger that decides whether the figure was ever real.
We are fluent, as a profession, in the machinery of benefits. We build benefit profiles, dependency maps, realisation plans; we assign owners and baselines and review cadences. What we are far less fluent in is the one test that actually matters — whether a hard-nosed finance function, asked to stake its own credibility on our numbers, would do so. It almost never would. And the uncomfortable argument of this piece is that finance is usually right not to, and that the programme’s habitual response to that disbelief makes the problem worse rather than better.
Two Ledgers, Two Languages
The disbelief is not, at root, about honesty or competence. It is structural. The programme and the finance function keep two different books, in two different languages, and those books were never designed to reconcile.
The programme’s book is a benefits ledger. It counts value in the units of the initiative: hours released, error rates reduced, handling times cut, cost per transaction lowered. These are real effects, often carefully evidenced. But they are counted at the point of the activity, on the assumption that a saving observed is a saving banked.
Finance keeps a budget ledger. It counts one thing only: money that a named person is accountable for, sitting in a cost centre, in a period, against a target. To finance, a benefit exists when — and only when — a line in someone’s budget goes down and stays down, and the person who owns that line has agreed that it should. Everything else is, in the CFO’s world, an assertion. And this is an unforgiving season in which to be selling assertions to a finance function: the reliability of the numbers a CFO attests to has rarely been under closer external scrutiny than it is now, and a figure they cannot defend has become a personal exposure rather than a professional embarrassment.
Between these two books lie the four faults through which almost every programme number leaks:
- Attribution. The saving lands, if it lands at all, in a cost centre other than the one that funded the programme — so the sponsor cannot point to it and the budget-holder who received it never asked for it.
- Timing. The benefit is claimed when the capability goes live; the budget can only fall a year or two later, in a planning round the programme will no longer be around to influence.
- Baseline drift. The “before” figure was estimated by the programme, for the business case, without finance’s agreement — so finance simply does not recognise the baseline the saving is measured against.
- The cashable line. Most of what a programme counts is capacity released, not cash removed. Freeing four days a month of a manager’s time is genuine, but unless a post disappears or a supplier invoice shrinks, nothing leaves the profit and loss account.
Each of these is enough on its own to justify a raised eyebrow. Together they mean that the honest translation of most eight-million-pound benefit cases into the language of the budget ledger is a much smaller, much later, and much more contested number. The CFO’s disbelief is not obstruction. It is arithmetic.
Walking the Eight Million Down
Let me be concrete, because this is where the argument is either earned or lost. Take that headline figure — eight million pounds of annual savings from a shared-services consolidation — and walk it down the way a sceptical finance analyst will, line by line.
| Claimed benefit | Amount | What survives contact with the budget ledger |
|---|---|---|
| Productivity / time released | £3.0m | Nil to the P&L — no posts removed, so the hours stay on the payroll |
| Cost avoidance vs a forecast rise | £2.0m | Nil recognised — the forecast increase was never in finance’s own plan |
| Already counted elsewhere | £1.5m | Nil — double-counted against a separate cost-reduction target |
| Genuinely cashable | £1.5m | Real — but lands in a different cost centre than the one that paid for the programme |
Of eight million pounds, perhaps one and a half is cashable, and even that arrives in the wrong pocket. This is not a fraudulent case; every line was defensible on its own terms. It is simply that only one of the four was ever going to change a budget. When the finance director asks why the cost base has not moved, this table — which the programme rarely constructs for itself — is the answer.
The instinct at this point is to defend the seven and a half. The productivity was real; the avoided cost would have hit us; the double-count is an accounting technicality. All true, and all beside the point. The programme is arguing in the benefits ledger while the decision is being made in the budget ledger. You cannot win an argument by speaking the wrong language more loudly.
The Objection From Both Sides
Two serious objections press on this argument from opposite directions, and a piece that ignored either would be rigging the case.
The first comes from finance itself, and it is the stronger of the two. If cashable benefit is all that counts, the CFO will say, then the remedy is simple: at the point of delivery, I cut the budget. Take the saving out at source, and the argument about belief disappears. This is the harvesting instinct, and within its own logic it is impeccable. But it fails in practice for a reason programmes see and finance often does not. A budget cut imposed without the operational capacity to absorb it does not realise a benefit; it defers a cost. The posts come out, the work does not, and eighteen months later the same money reappears as overtime, as contractors, as a service that has quietly degraded and must be repaired. A saving harvested before the capability is genuinely there is not a saving. It is a loan against next year, taken at a punitive rate.
The second objection comes from the programme side, and it is the one we tell ourselves. Finance, we say, is simply conservative — congenitally unwilling to credit value it cannot touch, and if we let the accountants set the test, no ambitious programme would ever be funded. There is a grain of truth here: an over-literal finance function can strangle genuine option value and long-horizon capability, and a CFO who recognises only next year’s cashable pound will underinvest in everything that matters the year after. But this is an argument for widening what finance counts, negotiated in advance — not for asking finance to believe a number it had no hand in setting. Conservatism met with better rhetoric hardens. Conservatism met with a baseline it agreed to, and a benefit booked in its own units, has nothing left to be conservative about.
A benefit that has not been subtracted from a real budget, owned by a named person whose target has changed, has not been realised. It has merely been described.
Making the Number Bankable
The move that resolves all of this is not analytical but organisational, and it inverts the usual sequence. Most programmes build the benefits case first and take it to finance for endorsement at the end. The disciplined programme does the opposite: it subordinates its benefits case to the finance ledger from the outset, so that the number it eventually reports is one finance has already agreed to own.
In practice that means a handful of things, none of them technically difficult and all of them culturally hard.
- Agree the baseline with finance before the programme starts, not after. The single most expensive omission in benefits management is a “before” figure the finance function never signed. Baseline the cost with them, in their cost centres, on their definitions — even where their definition gives you a smaller number than your own. A smaller baseline you can both defend beats a larger one only you believe.
- Name the budget-holder who will lose the money. Every cashable benefit must attach to a specific person whose budget will fall and whose target will be reset accordingly. If no one’s target changes, no benefit will be realised, however elegant the tracking. The test of a benefit owner is brutally simple: are they willing to have next year’s budget cut by the amount they are signing for?
- Promise only the cashable line to the profit and loss account. Keep counting productivity, cost avoidance, and capacity released — they matter, and they justify much of what a programme does. But present them in their own column, never summed into the headline the CFO will test against the cost base. Let the bankable number be small and true rather than large and disbelieved.
- Book the saving as a target reduction, timed to the planning round. A benefit that is not written into next year’s budget in the cycle when budgets are actually set will evaporate, because the organisation’s default is to spend to the number it holds. Land the reduction in the plan, or accept that it will not land at all.
None of this makes the benefits case bigger. It makes it bankable — which is the only property finance was ever asking for. The programme that does this stops arriving at the benefits review to be disbelieved, and starts arriving to confirm a number the finance director has already put in the plan.
The CFO Belongs in the Room Early
The reflex, when finance will not believe the numbers, is to treat the finance function as an obstacle to be managed — briefed carefully, brought onside late, handled. The better reading is the opposite. The CFO’s disbelief is not an obstacle to the benefits case; it is the discipline the benefits case was missing. The finance function is simply the first reader who applies, without sentiment, the test every programme’s numbers must eventually pass.
Which is why the practitioners who have made their peace with this do not keep finance at arm’s length until the review. They want the analyst who will one day walk the eight million down to be in the room while the case is still being built — precisely because that walk-down is going to happen either way. It can happen early, in private, when the number can still be made honest. Or it can happen late, in the steering group, when all that is left to do is watch it soften in the silence after a single question.