Home · Proof · Value measurement

How I Quantify Value in Finance Transformation

Most transformation business cases fail the same way: they claim benefits nobody can audit. This is the four-part taxonomy I use instead — with the arithmetic shown, from an enterprise programme across ten manufacturing entities.

Hard savingsMoney that stops leaving
Cycle timeElapsed days, auditable
Capacity releaseHours redeployed, not cut
Decision valueWhat the organisation can now see

There is a moment in every transformation conversation where someone asks: "So what did it actually save?" It is a fair question, and most answers are bad ones.

The bad answers fall into two camps. The first inflates — headcount reductions that never materialised, benefits double-counted across workstreams, savings that assume the organisation stops doing something it in fact still does. The second deflates — "it improved visibility," "it enabled better decisions" — true, unfalsifiable, and worthless in a business case.

What follows is the discipline I apply instead: four distinct value types, each measured differently, each honest about what it is. Where I have the underlying inputs, the arithmetic is shown. Where a figure is an estimate, it says so.

1

Capacity release

HOURS REDEPLOYED — NOT HEADCOUNT REMOVED

This is the value type most often misrepresented. Consultants translate hours into full-time equivalents and then into salary savings, which implies people leave. In practice they rarely do — and finance leaders know it, which is why that number gets discounted the moment it is presented.

The honest version measures the same hours but claims something different: capacity redeployed from data compilation to analysis. Same arithmetic, defensible conclusion.

The clearest example in my own work is the monthly consolidation of standard cost by product across ten manufacturing entities on different ERP systems. Every cycle, the same sequence ran by hand.

Standard cost consolidation — the manual cycle, before
Site controllers extracting and validating cost by product
10 controllers × ~2.5 hours each
~25 h
Business analyst consolidating submissions and loading to the ERP
reconciling total cost and cost by product against the reporting system
~3.5 h
Preparing and distributing the report to product managers and business leaders~3 h
Per monthly cycle≈ 31.5 hours
Annualised (12 cycles)≈ 378 hours ≈ 47 person-days

Basis: participant-reported effort for the recurring cycle, excluding error correction, restatements and the ad-hoc requests that followed each distribution. The figure is therefore conservative. It counts only work that disappeared — not work that became faster.

Before
  • Ten spreadsheets, ten formats, ten submission dates
  • Two reconciliation checks per entity — total cost, and cost by product
  • Consolidation and ERP load by hand at a single site
  • Errors surface downstream, after distribution
  • Analysts spend the cycle compiling, not analysing
After
  • Product-level standard cost held in the consolidation platform
  • Validation systematic, at the point of entry
  • No manual consolidation step, no re-keying
  • Exceptions handled, rather than files assembled
  • The same people, on analysis instead of assembly
Roughly 47 person-days a year stopped being spent on assembling data — in a process that produced no insight until the assembly was finished.
2

Cycle time

ELAPSED DAYS — THE MOST AUDITABLE NUMBER YOU HAVE

Cycle time is the value type I trust most, because it cannot be argued with. The close either finished on working day three or it did not. There is a calendar, and everyone was there.

The compression I led took a reporting cycle from two weeks to three days, standardised within three months and sustained thereafter. What matters for a business case is not the headline percentage but what the recovered days were used for.

Where the recovered time went

Two weeks of elapsed close is not two weeks of work — it is work interleaved with waiting: waiting for submissions, for reconciliations to clear, for questions to come back. Compressing the cycle removes the waiting, and the recovered days move to the front of the following month, where analysis actually influences decisions rather than explaining decisions already taken.

The team involved was small and senior: an analyst, an FP&A manager, and a controller or assistant controller. I have deliberately not converted their time into a monetary figure, because I do not have a reliable measure of what proportion of each working day the close consumed. Publishing an FTE saving on an assumed percentage would be exactly the kind of number this page argues against.

The harder version: Argentina

The same discipline applied under considerably worse conditions. In Argentina, a single legal entity contained three reporting entities — a manufacturing site, a sales entity and a headquarters entity — with the close running to roughly two weeks in an inflationary environment, on accounting that needed alignment to group standards.

The harmonisation ran three months from September, and by year-end the entity was on track with the standard group cycle. By March of the following year it was running a soft budget and forecast process alongside a close that finally left room for reconciliations and standard cost harmonisation — that is, for the work that prevents the next problem rather than explaining the last one.

A close cycle is not shortened by working faster. It is shortened by removing the waiting — and the test of whether it holds is whether it survives the next year-end.
3

Hard savings

MONEY THAT STOPS LEAVING THE COMPANY

Hard savings are rare in finance transformation, which is precisely why they deserve careful handling when they occur. Most of what a transformation produces is capacity and better decisions; genuine cash impact is the exception.

One occurred here, and the mechanism is worth understanding because it is structural rather than clever.

Why the old model broke

Sales commissions were calculated on a margin basis that did not incorporate the standard cost and manufacturing margin from the production sites. That was tolerable while input costs were stable. After 2021 it stopped being tolerable: inflation eroded manufacturing margin, standard costs rose, and real margin fell — while commissions continued to be paid against a measure that could not see any of it.

The company was, in effect, paying incentives on profitability that no longer existed.

Once consolidated manufacturing margin was incorporated into the incentive calculation, payouts recalibrated by roughly 10–15% in the largest growth centre. That is cash that stopped leaving the business — and it recurs, every cycle, for as long as the measure stays correct.

There was resistance, and it was rational: people were being measured on something harder. The resistance was also the point. An incentive that cannot see manufacturing margin will reliably produce sales that do not earn any.

This was not a cost-cutting exercise. It was an accuracy exercise — and accuracy happened to be worth double digits.

A note on what is not published here: the absolute value of that recalibration is a former employer's commercial information, and it is not mine to disclose. The percentage is what matters for the mechanism, and the mechanism is what transfers to another company.

4

Decision value

WHAT THE ORGANISATION CAN NOW SEE — AND THEREFORE DO

This is the value type that resists measurement, and the one that ultimately mattered most. It cannot be reduced to a single figure, so I describe it as a change in capability with observable consequences.

Before the transformation, product-level profitability data was effectively confidential — held centrally, released selectively, and by the time it reached the people who could act on it, the period was closed.

After, every finance director and controller could pull the data from the consolidation system directly. That single change turned a reporting function into a business partnering function, because a controller who can answer a sales director's question in the meeting is a different professional from one who has to request the data and reply next week.

Three consequences that followed

R&D refocused

Development effort moved toward the products that were genuinely profitable at consolidated margin — not the ones that looked profitable at sales margin alone.

Monthly product action

Product managers began acting on the lowest-margin lines every month, rather than discovering the problem at year-end review.

Controllers advising sales

Finance directors and controllers moved from producing reports to providing insight to sales directors — with the same data, in the same room, at the same time.

None of these three appears in a savings line. All three are the reason the framework outlived my tenure and became the standard way management measures and incentivises performance globally — the outcome recognised with three CFO Excellence Awards.

The most durable value in a finance transformation is rarely the money it saves. It is what the organisation becomes able to see — because that changes what it does, every month, long after the programme closes.

Methodology and honest limits

Any number on this page should be readable by a sceptical CFO without generating a follow-up question that has no answer. Here is the basis for each.

What is measuredElapsed cycle times, entity counts, and the recurring task structure of the consolidation process. These are observable facts of the process as it ran.
What is estimatedHours per task, based on participant-reported effort for a normal cycle. Estimates exclude error correction, restatements and downstream ad-hoc requests — making the totals conservative.
What is deliberately absentFTE conversions and salary-based savings, since I hold no reliable measure of what share of each working day the close consumed. Absolute commercial figures belonging to a former employer are also not disclosed.
What does not transferThese outcomes came from one organisation, in one industry, over several years. The taxonomy transfers. The numbers do not — they are illustration, not benchmark.

The reason to be this careful is practical rather than moral. A business case survives contact with a CFO only if every figure in it can be traced to how it was derived. Numbers that cannot be defended do not merely get discounted — they discredit the ones that could have been.

← Back to Proof