I had coffee last month with a founder doing $6M in revenue. She told me her bookkeeper “basically does CFO stuff.” When I asked what that meant, she said, “You know, the money stuff.”
She's not alone. Most founders I talk to think a CFO is just a senior accountant — someone who makes sure the books are right and the taxes get filed. That's like saying a product manager is just a project manager with a fancier title. The roles share a surface-level resemblance, but the actual work is completely different.
Here's what a CFO actually does, broken down into five distinct jobs. Understanding these will help you figure out whether you need one now, later, or whether technology can fill some of the gaps in the meantime.
Job 1: Cash Flow Architect
This is the one most founders eventually feel in their gut. A CFO doesn't just track where money went — they engineer where money goes.
That means building cash flow models that project 13 weeks into the future. It means knowing that even though you just closed a $200K deal, the payment terms are Net 60, and payroll hits in 14 days. It means structuring vendor payments, credit lines, and receivables so you never end up asset-rich and cash-poor.
I've seen a $4M services business that was profitable every single quarter for two years — and still had to take on a short-term loan to cover a gap between when clients paid and when contractors needed to be paid. A bookkeeper records that gap. A CFO prevents it.
Job 2: Strategic Financial Planner
This is the “should we hire three people or five?” question. The “do we open a second location?” question. The “can we afford to drop our lowest-margin product line?” question.
A CFO builds the financial models that turn gut-feel decisions into math. They'll tell you that hiring five people only works if you maintain 72% gross margin on the new revenue those hires generate. They'll show you that opening a second location has a 14-month payback period — but only if you hit $38K/month in revenue by month four.
Without this, founders make one of two mistakes: they play it too safe (leaving growth on the table) or they swing too aggressively (burning cash on bets that don't pencil). Both cost money.
Job 3: Risk Manager
This one is invisible until it isn't. A CFO watches for things that could kill the business — not the obvious stuff like “what if revenue drops,” but the second-order risks.
What happens if your biggest client (30% of revenue) churns? What's your exposure if interest rates jump another 200 basis points and your line of credit reprices? What if a key vendor goes bankrupt and you need to find a replacement at 20% higher cost?
A good CFO runs these scenarios quarterly. They maintain a “what if” model that quantifies the impact of each risk and identifies the trip wires — the early indicators that a risk is materializing. Most $3M–$10M businesses have zero risk modeling. They find out about risks when they become problems.
Job 4: Capital Strategist
When do you raise? How much? At what terms? Should you take on debt instead of equity? Is revenue-based financing a better fit than a traditional line of credit?
These are capital structure decisions, and they have enormous long-term consequences. I've watched founders give up 25% of their company for $500K because they didn't realize they could get a $300K SBA loan at 7.5% that would have covered the same gap. That's a million-dollar mistake that takes five minutes on a spreadsheet to avoid — if someone knows to run the numbers.
A CFO handles investor relations, manages banking relationships, and makes sure the company's capital structure matches its growth plan. For startups considering their next move, this is often the most valuable role a CFO plays. (We wrote a deeper dive on the raise vs. bootstrap decision if you're at that crossroads.)
Job 5: Accountability Partner for the Numbers
This one is underrated. A CFO creates the financial rhythm of the business: monthly closes, quarterly reviews, annual planning cycles, KPI dashboards that the leadership team actually looks at.
They're the person who sits in the Monday meeting and says, “We said we'd hit $420K in revenue this month. We're at $380K with six business days left. Here's what needs to happen.” They turn financial goals into operational actions.
Most small businesses don't close their books until 3–4 weeks after month-end. By then, the data is a historical artifact, not a management tool. A CFO compresses that cycle and makes the numbers actionable.
So When Do You Actually Need One?
There's no magic revenue number. But here are the signals I've seen that mean it's time:
- You're making financial decisions on instinct because pulling together the data takes too long or feels unreliable
- You've been surprised by cash — either running low when you expected to be fine, or sitting on more than you realized and missing investment opportunities
- You're approaching a capital event — fundraise, acquisition, major debt facility — and nobody on your team has done one before
- Your accountant or bookkeeper is doing their best, but they're answering “what happened” and nobody is answering “what should we do about it”
- You're past $2M in revenue and the financial complexity is outgrowing your spreadsheets
The Options Between “Nothing” and “Full-Time CFO”
A full-time CFO costs $150K–$300K. For most businesses under $10M, that's not realistic. But you're not stuck choosing between that and nothing:
- Fractional CFOs give you 10–20 hours per month of senior financial leadership at $5K–$15K/month. Good for businesses in the $3M–$20M range that need strategic guidance but not a full-time seat.
- AI-powered financial intelligence handles the monitoring, anomaly detection, and real-time visibility that a human CFO would otherwise do manually. It won't negotiate your next fundraise, but it will tell you at 2 AM that your burn rate just spiked 18% because a vendor raised prices.
- The combination is where most smart operators are landing: AI for the continuous data layer, a fractional CFO for the quarterly strategy and big decisions. You get 80% of a full-time CFO at 20% of the cost.
The worst option is the most common one: doing nothing and hoping the numbers work out. They usually do — until they don't.
