Insight · The finance seat

11 Pressures Mid-Market CFOs Are Quietly Managing

None of these is new on its own. What is new is that they all arrived together — and that the job has grown faster than the authority that comes with it.

For chief financial officers All sectors 13 min read

In brief

Eleven pressures that mid-market CFOs are carrying across transformation, AI, regulation, talent and the board. They are structural rather than personal.
The common thread is an asymmetry: the CFO is accountable for outcomes the CFO did not design. That is why working harder or hiring better does not relieve them.
The answer is not better personal management. It is a change that puts authority where the accountability already sits. Most firms have not had that conversation.

Mid-market finance chiefs are carrying more than at any point in recent memory. Not because any single pressure is new, but because they have all arrived at once.

Transformation programmes running below the approved case. AI spend rising without anyone governing the whole of it. Compliance costs that have become permanent. Boards asking for analysis the data cannot easily produce. Good commercial finance people harder to keep than ever. The chair is bearing more weight than it was built for.

What is uncomfortable, in private, is how much of this gets managed without ever being named at board level. Each pressure looks individually like a problem to solve. Together they describe something else: a mismatch between what the job is now expected to deliver and what the job actually controls.

This is written for the finance chief who is quietly managing all of it and would like confirmation that it is structural rather than personal. Each of the eleven is recoverable. Almost none of them is recoverable through individual effort, because the changes that would relieve them sit at executive team level rather than inside finance.

01

The gap between the approved case and what actually landed is widening

Investments get approved against optimistic cases. Finance governs delivery against the approved numbers. Afterwards the returns are partial. The variance gets reported. The board notices.

Five years ago this was manageable from the finance seat. The cases were smaller, the cycles shorter, and the variance usually traced back to something specific you could influence.

Now it is structural. It comes from changes to the way people work that never happened, from adoption that fell short, from scope that grew without anyone recording it. You can report the variance precisely. You cannot fix it from where you sit, because the causes are in other people’s functions.

02

AI is being approved one case at a time, with nobody governing the whole

Each use case arrives with its own business case. Marketing. Service. Finance. Operations. Each is approved on its own merits. You cost each one. IT designs each one. Nobody sees the total.

Meanwhile the aggregate is climbing steadily while the returns are uneven. If someone asked you today for a single page showing total AI spend by use case with the return against each, you could not produce it quickly — not because you have not thought about it, but because nothing was set up to capture it.

Boards are starting to ask for exactly that page. The work of assembling it usually reveals more than the page itself does.

03

Forecast accuracy suffers during change, and nobody separates the two

In steady state, forecast accuracy is one of the visible measures of whether finance is any good. The variance gets reviewed, the trend matters, and boards notice when it moves.

During major change, variance goes up structurally. Several things move at once. Customers behave differently during migration. Costs spike at cutover. Benefits arrive unevenly rather than in a line.

You know the variance is caused by the transformation. The board sees the variance. What almost never exists is an agreed distinction between forecast accuracy during change and forecast accuracy in steady state — so you defend the number without the framing that would make the defence easy.

The finance chief is increasingly accountable for outcomes the finance chief did not design. Almost everything else on this list follows from that.

04

Hiring is deferred but the work did not go anywhere

A hiring freeze is far easier to defend at board level than a complicated investment case. So headcount gets deferred through cost discipline. The work those people would have done gets absorbed — by existing teams working longer, by consultants, by programmes quietly carrying extra scope.

The total cost of the work has not fallen. Its visibility has moved. You know this. The board sees the headline discipline on headcount. The money is the same or more, sitting in different lines.

This is one of the quietly uncomfortable ones, because you are managing the appearance of cost control while the underlying economics have not changed. The fix is to model the true cost of the work rather than the cost of the people, and very few firms do that consistently.

05

Consultancy spend has quietly become a permanent line

Year on year, spend with consultancies and integrators is stable or growing. You budget it. The board reviews it. It stopped being the cost of change some time ago. It is now simply how the business operates.

This is what happens when the internal capability that would have replaced it never got commissioned, because the case for hiring kept losing to the case for another engagement.

You see it as structural. The board often still sees it as project spend. The conversation about bringing the work in-house is genuinely hard, because it needs multi-year investment judged against this year’s cost discipline. Most firms have never had it explicitly.

Where do you sit?

Recognising that these are structural rather than personal is the first step.

A free Executive AI Opportunity Assessment. Forty-five minutes, online, with your leadership team. We arrive with a model of how firms in your sector make money and where AI has moved those numbers — and you tell us where it is wrong.

Book a free assessment

06

Software gets bought on credit cards and nobody sees the total

Marketing buys an analytics tool. Customer success buys an engagement platform. Operations buys something for workflow. Each individual purchase sits below whatever threshold would have triggered procurement.

The annual total is not small. You find out when the renewals stack up, when an integration breaks, or when someone finally proposes consolidating vendors. Nothing was ever set up to watch it, because no single purchase justified setting anything up.

The cost is structural and growing, and you are reconstructing the picture from invoices. The fix is someone owning the whole portfolio with monthly visibility — which is work you are rarely funded to do, and which nobody notices until it is done.

How the job has changed

What	Five years ago	Now
Investment	Allocating capital against a case	Governing a portfolio and attributing returns
Forecasting	Plan against actual, in steady state	Plan against actual, mid-transformation
Headcount	A direct cost line	Hidden inside consultancy and programme budgets
Software spend	Occasional procurement events	Continuous, across every function
Compliance	Periodic projects	A permanent line in the accounts
AI returns	Not yet a finance question	Expected to stand up to scrutiny
How the business runs	Consulted on the cost impact	Accountable across every function
The board	Annual budget, quarterly review	Continuous challenge, evidence expected

07

Compliance costs have grown without anyone asking what else they could buy

Consumer duty. Operational resilience. AI governance. Money laundering. Sustainability reporting. The compliance line has grown materially over five years and is now permanent.

What has not grown is any commercial thinking about it. Doing compliance well is increasingly a genuine advantage — but the same pound is rarely designed to buy both defensibility and advantage. You see a cost. The commercial team sees friction. Nobody sees the opportunity.

The fix is reframing rather than more money. The same spend, designed with both lenses, returns very differently. Few finance teams have built that into how the case gets written.

08

Good commercial finance people are harder to keep than they have ever been

Senior planning and analysis people, business partners, commercial finance leads. The market is tight, pay expectations have risen sharply, and people stay for less time. The ones who do become available are often poached back within a year.

So you make do with a smaller and less experienced team than the workload needs. The visible response is leaning harder on the people you have and accepting more external help. The invisible response is that the capacity to do the structural work — the work that would relieve half of this list — disappears.

The fix is genuinely multi-year: apprenticeships, developing people internally, designing roles worth staying in. All of it competes with this year’s cost discipline, which is why it keeps getting deferred.

For the finance chief

Three questions before the next planning cycle

Across everything in flight, what is the total gap between what was approved and what was delivered? If it needs assembling, nothing was built to show you.
Of the pressures on your role, which sit inside your formal authority and which sit outside it? If most sit outside, that is the whole problem in one answer.
What would relieve these structurally, independent of how hard you work? If the answer is “work harder” or “hire better”, keep going until it is not.

09

The board wants better analysis than the data can support

Cohort views. Attribution. Scenarios. Variance in something close to real time. All reasonable asks. None of them are things the data was set up to answer.

So the team assembles it by hand. The papers look polished and rest on preparation effort the board never sees. The gap between what gets asked for and what can be produced sustainably keeps widening.

The fix is real investment in the finance data foundations, which competes with everything else and keeps losing, because the current state is acceptable right up until it is not.

10

The job grew. The authority did not.

Strategic finance. Governing transformation. AI returns. Capital allocation into categories that did not exist five years ago. The responsibilities have expanded materially.

Authority over how the business actually runs has not expanded alongside. Those decisions still sit with the sales director, the operations director, the customer director, the technology director. You contribute. You do not decide. And you are held accountable for the results.

This is the source of most of the rest of the list. Until the authority is reconciled with the accountability, every pressure here gets managed personally rather than fixed structurally.

11

You end up defending outcomes you did not design

The board asks why the benefits have not appeared. The programme team can defend the delivery. Each function can defend its own numbers. The question about why the whole thing has not produced what was promised lands on your desk.

This is the asymmetry made visible. You defend cumulative outcomes you did not design, did not write, and had no authority to change. The defence involves explaining decisions made by peers, without naming them, while absorbing accountability for a gap with causes across the whole business.

It is wearing in a way that is difficult to raise, because raising it sounds like complaining about colleagues. So it gets carried quietly, which is exactly why it never gets fixed.

What this means

None of these eleven are personal. None are about capability. They describe a job that has grown faster than everything around it, and a set of pressures that compound when each one is handled alone.

The thread through all of them is the gap between accountability and authority. Sales designs how selling works. Operations designs how delivery works. Technology designs the systems. Finance sits inside the consequences of all three, holds the cost discipline across them, and explains the combined result to the board.

The economics of that are quietly awful. Managing structural pressure personally consumes exactly the capacity that would be needed to build the fixes. That is the trap, and it is a hard one to see from inside because every individual month looks survivable.

Getting out of it is structural rather than personal. It usually means the executive team agreeing that whoever has to defend the outcome should have had a say in designing it — which, when it comes to AI in particular, means settling what the work will become before anyone signs for the technology. Change the work before you change the technology, and the finance seat stops defending decisions it never made.

The next step

Are the pressures on your seat structural, or personal?

Forty-five minutes with your leadership team, online, at no cost. We bring a working model of how firms in your sector make money and where AI has moved those numbers. You tell us where it is wrong. Then we tell you what we think — including “don’t invest”, if that is the honest answer.