Underinvested Finance: The Hidden Adjustment in Your EBITDA

Why the cheapest finance function in the data room deserves the hardest look, and why the savings case grows or stalls on how the transition is run.

Finance cost as a percentage of revenue is the number everyone quotes in diligence. On its own it is the least useful one. The same ratio can describe an efficient function or a neglected one, and four signals sit behind it that change what you do with the investment case.

The worked example below uses one illustrative company: US400Mofrevenue, an832M), and finance running at 1.5% of revenue (US$6.0M) with 45 FTEs.

Four signals behind the ratio

1. Underwrite-able cost-out, not just "synergy." Taking finance from 1.5% to 1.0% of revenue releases 50 basis points and lifts margin from 8.0% to 8.5%. That is US$2.0M, or about 6% more EBITDA, and it does not depend on top-line growth. A revenue synergy asks customers to behave as the model assumes. This cost line is within your control, which makes it one of the more certain items for a 100-day plan.

2. A signal of operational maturity, and of hidden problems. High cost-to-serve tends to travel with a manual close, fragmented ERP instances, no shared services and weak data lineage. That suggests reported EBITDA may be weak too, so pressure-test add-backs, manual journal entries and revenue recognition before trusting the CIM. The reverse also holds. A function that looks cheap because it skips what a buyer will later expect carries a hidden adjustment: the cost of fixing it.

3. Scalability, or the lack of it. If cost-to-serve is high, volumes are flat and headcount rises in a straight line, the function will not scale. That counts against a growth thesis and in favour of buy-and-build, where each bolt-on's duplicate finance cost can be centralised into a shared services centre.

4. Technology spend versus spend waste. In my experience, stronger finance functions spend more on technology (RPA, e-invoicing, AI-based anomaly detection in AP and close) and far less on labour. A company at 0.9% of revenue with little automation has clear headroom. One that is already automated and still at 1.2% has an organisation-design or complexity problem, which is slower and harder to underwrite.

An illustrative EBITDA bridge

Of the US6.0M, aboutUS5.0M is people and US1.0Misnon − peoplecostsuchasERPruncostandtools.AssumeUS100K per transactional FTE, fully loaded, and about 27 of the 45 roles transactional.

The levers act on different work, so they do not overlap: of the 27 transactional roles, automation removes 5, shared services removes 6 duplicated across entities, and outsourcing takes 10 of the remaining 16.

The result is finance at US4.0M(1.03.0M to US$4.0M, with payback inside a typical hold period. If only 75% of the saving is delivered, the uplift falls to about 4.7%.

The savings case is not fixed on day one

I set up the global finance shared services centre in Bangalore for a US-headquartered technology manufacturer. It was the company's first attempt at consolidating finance globally in one place. The approved business case was modest: about 30 people, focused on low-cost transactional work, accounts payable and accounts receivable.

That was a sensible scope for a first attempt, but it capped the value. The larger opportunity sat in the work closer to the numbers, and the business teams had good reason to guard it. What changed their position was assurance. I built the case for expansion around a transition plan that showed, phase by phase, how each team's interests and the customer's would be protected, and I made sure senior stakeholders heard that message consistently. Review systems and KPIs then gave them evidence instead of promises.

The scope grew in stages. The centre took on fixed assets accounting, which is complex in a capital-intensive group, then revenue accounting and balance sheet reconciliations. Tax, payroll, FP&A and costing followed. Over about eighteen months the team grew from 30 to more than 120 people.

The lesson for an investor is that the size of the finance opportunity is set by how much of the work the business will trust to move. A plan that starts with what the business will release, and earns the rest through visible control, tends to beat one that assumes the full scope on day one. Where the scope ends up matters as well. Revenue accounting, fixed assets and reconciliations sit directly behind reported EBITDA, so a centre that runs them well improves the quality of earnings as well as the cost.

Where the case gets weaker on inspection

• Overlap between levers. Reconcile automation, shared services and outsourcing at process level before adding them up.
• Dependencies. The ERP saving sets the pace for anything that relies on one instance.
• Dis-synergies. Retention, parallel running and a temporary dip in close quality belong in the bridge.
• A low ratio that reflects control debt. If the function is cheap because it skips what a buyer expects, the fix adds cost before it removes any.

What this means for the investment case

Ask for finance cost-to-serve by process, not in total, and use it to test the quality of earnings as well as the size of the opportunity. Separate the levers, phase them against the hold period, and net off the cost to achieve. Then ask a question most models skip: what does the transition plan say about protecting the business while the work moves? Where that answer is weak, the scope, and the savings, are likely to stall.

How I approach it. I start from headcount, volume and system data by process, build a range with the assumptions written down, and test it against close and control quality. The output is a bridge an investment committee can challenge line by line.

Axiom Finance Advisory works with deal teams and portfolio companies on finance operations diligence and transformation. All figures in the worked example are illustrative and are not benchmarks.