Why the finance function in most small businesses is either invisible or overwhelmed — and what AI actually does about it.
This is the most common finance setup in small businesses — and nobody talks about it. One person handling everything. Not because the company is poorly run. Because that's how it works at this size. And for most of those companies, it has always worked — until it doesn't.
I've seen this ratio in more companies than I can count. Sometimes it's 30 people and €8M. Sometimes it's 80 people and €22M. The exact numbers change — the structural problem doesn't. The finance function is undersized relative to what the business actually needs from it.
The company isn't making a mistake. It's making a rational decision: a full finance team is expensive, and at this scale, one capable person can handle the basics. Invoicing. Payroll. Monthly close. Tax filings. That's the job description. It's also a full-time job. So the basics are what gets done.
The finance function in a 50-person company is usually one of two things: invisible or overwhelmed. Sometimes both at once, depending on the month.
"Undervalued" doesn't mean the accountant is unappreciated. It means finance is treated as a compliance function, not a strategic one. The numbers get reported after the fact. Nobody is running cash flow forecasts three months out. Budget variances get explained in retrospect. Nobody is asking "why did operating costs jump 12% last quarter" until the end of the quarter.
"In a 50-person company with one accountant, finance is almost always reactive. Not because the person isn't capable — but because there's no time left to be proactive."
Here is a realistic breakdown of a typical week for a solo accountant in a growing SMB. I've seen variations of this across dozens of companies. The proportions shift slightly — the pattern doesn't.
That 5% for forecasting and analysis is not a personal failure. It's a structural one. When one person is responsible for keeping the operational finance running, there is no slack in the system for anything strategic. The month-end close doesn't move. Invoices don't wait. Payroll runs on a deadline.
The company grows. Revenue goes from €8M to €15M. Headcount goes from 30 to 50. Transaction volume doubles. The number of people in finance stays at one. Because hiring a second finance person means admitting the function needs real investment — and the business has never had to make that call before.
So the accountant absorbs it. Works longer hours. Gets faster at the routine tasks. Stops taking vacation in March and September. Becomes the single point of failure for anything finance-related. And still doesn't have time for the forecasting, the scenario analysis, the proactive cash management that would actually help the CEO make better decisions.
"The company's finance team isn't failing. It's succeeding at the wrong things — because those are the only things there's time for."
The answer most people reach for is: hire a second person. That's not always wrong. But it's also not always the right first move — especially when what's broken is the time allocation, not the headcount number itself.
What AI does is shift where the time goes. Not by replacing the accountant — by automating the parts of the job that are high-volume, low-judgment, and deeply repetitive. Data pulling. Report formatting. Variance flagging. Recurring reconciliations. The work that fills the week without adding insight.
The 10× figure is an illustrative estimate of the capacity effect, not a benchmark — a rough approximation of what happens when you remove the repetitive process layer from a one-person finance function. The person doesn't become ten people. The function starts operating like it has ten people's worth of coverage on the routine work — which frees the actual person to do the work that requires judgment.
This is a composite pattern, based on what I've seen repeat across multiple companies throughout my career — not a single specific case. The first thing that changes is not the quality of the reports — it's the timing. Reports that used to arrive on the 10th of the month start arriving on the 2nd. The CEO starts making decisions with current data instead of data from three weeks ago.
The second thing that changes is what the accountant does with recovered time. In most cases I've seen, it goes toward things the business genuinely needed but was never getting: rolling cash flow projections, scenario planning before a major hire or investment, proactive flagging of margin compression before it shows up in the P&L conversation.
The finance function doesn't get bigger. It gets better at the thing it's supposed to be — a system that helps the business make decisions, not just a system that records what already happened.
"AI doesn't replace the person in finance. It replaces the version of that person who spent 95% of their week on tasks that don't require a person at all."
The invoice table, AI-generated emails, and the 5-step pipeline — all visualized. Code on GitHub.
Post #4 moves to a different process entirely — cash visibility instead of invoicing — and covers the specific tasks that, in my experience, make sense to automate first, and why the order matters more than the tools themselves.
How many people are in your finance team right now?
No spam. One post per week. Unsubscribe any time.