A controller owns the accuracy of your past: clean books, closed months, compliant reports. A CFO owns the shape of your future: forecasts, fundraising, and capital allocation. Most startups need controller-level accuracy first, but that no longer means hiring a controller first, because software and AI agents now handle most of that execution.
That distinction matters because the two roles are priced very differently, and hiring the wrong one first is one of the more expensive mistakes an early-stage company can make. This guide covers what each role actually does, what each costs in 2026, the revenue thresholds most advisors use, and why the standard hiring timeline deserves a second look.
What a financial controller actually does
A controller is the most senior accountant in the company. Their mandate is accuracy: every transaction recorded correctly, every month closed on time, every report defensible under GAAP. They manage the accounting team, enforce internal controls, and own the month-end close.
The controller's work faces backward by design. Payroll, tax filings, audit preparation, accounts payable and receivable, and reconciliations all describe money that has already moved. That is not a criticism. Investors, lenders, and acquirers all price a company partly on whether its historical numbers hold up under scrutiny, and the controller is the person who makes sure they do.
At larger companies the controller reports to the CFO and acts as their operational counterpart. At smaller companies without a CFO, the controller often reports directly to the CEO and becomes the de facto head of finance, whether or not the strategic side of the job actually gets done.
What a CFO actually does
A CFO is a strategic officer, usually the second most consequential financial voice in the company after the CEO. Their mandate is the future: building the revenue forecast, managing burn rate against runway, running fundraising processes, negotiating with investors, and pressure-testing the plan through scenario planning before the board does.
Where the controller answers "what happened," the CFO answers "what should we do next." That includes pricing strategy, hiring pace, capital structure, and the awkward conversations about which parts of the business are quietly losing money. A good CFO also represents the company externally to banks, investors, and acquirers.
Most startups cannot justify this as a full-time salary before Series B, which is why the fractional CFO for startups model exists. A fractional engagement buys senior judgment for a few days a month at a fraction of the cost, and our breakdown of fractional CFO rates covers what founders should expect to pay at each stage.
Controller vs CFO at a glance
The comparison below reflects role definitions and compensation data published across 2024 and 2025, including Salary.com figures cited in 2025 reporting and BILL's 2024 salary analysis. Startup compensation typically lands below the large-company medians and adds equity.
| Controller | CFO | |
|---|---|---|
| Mandate | Accuracy of the past: books, close, compliance, controls | Strategy for the future: forecasting, fundraising, capital allocation |
| Reports to | CFO, or the CEO when no CFO exists | CEO and the board |
| Salary range (2024 to 2025 data) | $110K to $180K at most private companies (BILL, 2024); Salary.com median base near $253K skews to large enterprises | $150K to $300K at most private companies (BILL, 2024); Salary.com median base near $436K skews to large enterprises |
| Fractional cost | Roughly $2K to $8K per month | Roughly $3K to $12K per month |
| When a startup needs the function | The moment transactions exist, so from day one | Seriously from your first priced round onward |
| When a startup needs the full-time hire | Often never before $10M revenue, and later if execution is automated | Typically $30M to $50M revenue, or Series B and beyond |
Two things stand out in that table. The gap between needing the function and needing the hire is enormous for both roles, and the controller function starts on day one while the CFO function can wait months, but neither justifies a six-figure salary at seed stage.
The standard hiring timeline, and where it comes from
Most advisory content converges on the same revenue-based ladder. Under $1M in revenue, a bookkeeper or outsourced accounting service is enough. Between $1M and $10M, add a controller, part-time at first. Somewhere between $10M and $50M, bring in a fractional CFO, then convert to a full-time CFO once revenue, fundraising complexity, or M&A activity demands it.
This ladder is sensible as far as it goes, and it correctly implies that the controller usually comes first. Books that cannot survive due diligence will sink a fundraise faster than a mediocre forecast will. Accuracy is the foundation; strategy built on wrong numbers is just confident fiction.
But the ladder carries a hidden assumption: that the work inside each role is fixed, so the only question is when you can afford the person. That assumption made sense when every reconciliation, accrual, and close checklist required human hands. It no longer describes how the work actually gets done.
The real question: how much of each role is execution?
Break the controller role into its parts and most of it is execution: categorizing transactions, reconciling accounts, chasing receivables, processing bills, running the close checklist, producing the reporting package. The judgment layer, deciding accounting policy, handling a genuinely novel revenue arrangement, or preparing for a first audit, is real but thin at the pre-seed through Series A stage.
The CFO role splits the opposite way. Building the model is execution, but deciding what the model implies about hiring pace, pricing, and raise timing is judgment, and judgment is most of the job. That asymmetry is the key to sequencing, because execution is exactly what software and AI agents now absorb.
This is the premise Futureproof is built on. Our AI finance team gives startups six agents for $1,000 per month flat: Vic handles bookkeeping, Remi runs accounts receivable, Theo manages accounts payable, Margo builds forecasts and FP&A, Hugo covers RevOps, and Nia prepares investor reporting and board decks. Together they cover the execution layer of both roles, which is most of a controller job and the mechanical half of a CFO job.
Once the execution layer is handled, the first-hire question inverts. You no longer need a $150K controller to produce accurate books; the agents produce them, with variance analysis flagging anything that drifts from plan. What you cannot automate is senior judgment about the future, which argues for buying fractional CFO time as your first human finance spend rather than a full-time controller. Founders stuck manually stitching their stack together face a different version of the same problem, which we cover in stop being the human API for your finance stack.
When you still need the human controller
Automation covers the common cases, not all cases. Some situations still call for experienced controller judgment, whether fractional, outsourced, or eventually in-house.
Companies heading into a first financial audit benefit from someone who has been through one and knows what auditors will pull. The same goes for multi-entity consolidation across countries, inventory-heavy businesses with complex cost accounting, and revenue arrangements unusual enough that recognition policy is genuinely debatable. Regulated industries add their own layer of compliance work that deserves human ownership.
The honest framing is that these are triggers, not defaults. A SaaS startup between pre-seed and Series A with standard subscription revenue usually hits none of them, and can wait until one appears before paying controller-level salaries.
A sequencing framework for founders
For a typical SaaS startup, the order that follows from all of the above looks like this. First, get the execution layer handled from day one through software and AI agents, so the books are clean before there is much to clean. If you want a quick read on how much time that buys you, the startup runway calculator shows how a $1,000 monthly finance cost compares with a first finance salary.
Second, add fractional CFO judgment around your first priced round, a few hours a month to own the forecast narrative, sanity-check unit economics, and prepare you for investor questions. Our guide to the best fractional CFO companies compares the main options. Third, hire a full-time finance lead only when a concrete trigger demands it: an audit, an acquisition, multi-entity complexity, or a board that wants a dedicated finance executive.
Notice what this sequencing avoids. It skips the stage where a startup pays a full-time controller to spend forty hours a week on work software does continuously, and it skips the stage where a founder serves as an untrained CFO because all the finance budget went to bookkeeping labor. When you do make a first finance hire, there is a strong case that your first hire should be a generalist who can direct the agents rather than a specialist who competes with them.
The controller vs CFO question, asked properly, is not about which title to recruit. It is about buying accuracy and judgment in the cheapest reliable form each one comes in, at the moment you actually need it. For accuracy, that form is increasingly software. For judgment, it is a senior human on fractional terms until scale justifies more. If you want the accuracy layer handled this month, you can start with Futureproof and put the first human dollars toward judgment instead.
Frequently asked questions
Is a CFO higher than a controller?
Yes. The CFO is a C-suite executive who owns the entire finance function and reports to the CEO and board. The controller is a senior manager or director who owns accounting specifically and reports to the CFO where one exists. In companies without a CFO, the controller is often the top finance person by default, but the scope of the role does not expand to match the vacancy on its own.
At what revenue should a startup hire a controller?
Conventional guidance says part-time controller help from about $1M in revenue and a full-time controller by $10M. In practice the trigger is complexity rather than revenue: audits, multi-entity structures, inventory, or unusual revenue recognition. A SaaS startup with standard subscription billing and automated bookkeeping can push the full-time hire well past the traditional threshold.
Can a controller become a CFO?
Often, but the promotion is a change of job, not a bigger version of the same one. Controllers are trained for accuracy and compliance; CFOs are judged on forecasting, capital strategy, and investor work. Controllers who make the jump usually build FP&A and fundraising exposure first, and boards frequently prefer to hire a proven CFO externally when a raise or exit is near.
Does a pre-seed startup need either role?
It needs the output of both, not the payroll of either. Accurate books matter from the first transaction, and a basic forecast matters from the first hire, but both can come from software, AI agents, and a few hours of fractional advice. The first full-time finance salary is usually justified somewhere between Series A and Series B, and the strongest early candidates for that spend are judgment roles, not execution roles.
What does a fractional controller or fractional CFO cost?
Fractional controllers generally run $2,000 to $8,000 per month depending on transaction volume and complexity. Fractional CFOs run roughly $3,000 to $12,000 per month, or $150 to $450 per hour, scaling with stage and scope. Both are far below the loaded cost of full-time equivalents, which is why the fractional model dominates the pre-seed through Series A market.



