AI in the Office of Finance does not cut headcount
Buy AI in the Office of Finance to cut staff, and you lose both: the saving and the edge. Yet this arithmetic shows up early in almost every project. We bring in AI, then we can cut two roles. That is what the business case says, and that is how it gets approved. I have seen this arithmetic many times, across projects and in board roles. It almost never works out.
The problem is not weak AI. It is that the work meant to disappear does not disappear. It only moves.
What AI actually takes on
AI is good at producing routine. Reports, reconciliations, a first draft of a forecast. That work was measurable, which is why it ended up in the savings plan. What AI does not take on is the judgment of whether the result is right and whether you can stand behind it.
Where the work goes
Where someone used to gather the numbers, someone now checks what the model puts out. Whoever feeds the AI has to keep the data clean. Whoever uses the result has to defend it in front of the board. Three new tasks replace the one old one: maintain the data, control the models, own the results. None of them is easier than what falls away.
Why the savings plan rarely works in the mid-market
In the mid-market the teams are small to begin with. The two roles the business case cuts often hold the knowledge of the data and the quirks of the business. Remove them, and you remove exactly what the AI needs to produce usable numbers. The savings plan saws off the branch the impact sits on.
The better business case
Bring AI into the Office of Finance to save heads, and you are thinking too short; in the end you get neither the saving nor the impact. Bring it in to steer faster and with more confidence, and you get both: better decisions, and a team finally working on the right questions. AI is a lever for steering. As a cost-cutting lever it is too good, and too expensive.
Whether your business case aims at savings or at steering shows within ten minutes. If you are unsure where yours sits, let us talk.