At some point every founder hears the same advice: "You need to hire someone for finance." What that advice almost never specifies is who. A controller and a CFO are both senior finance roles, both expensive, and both routinely described as "the finance person." They are not the same job, and hiring the wrong one first is one of the more common and more costly sequencing mistakes venture-backed startups make.
This guide covers what a controller actually does, what a CFO actually does, what each costs at typical market rates, and the hiring order that makes sense for a venture-backed startup, which is different from the order most generalist articles recommend.
The Short Answer
A controller is the top of your accounting function. They own the monthly close, the accuracy of your financial statements, internal controls, and compliance. They look backward and inward: what happened, is it recorded correctly, and can we prove it.
A CFO is the top of your finance function. They own strategy, fundraising, forecasting, capital allocation, and the relationship with your board and investors. They look forward and outward: what happens next, what should we do about it, and how do we fund it.
The controller makes sure the numbers are right. The CFO decides what to do because of them. A startup eventually needs both capabilities. The question this article answers is which one you need first, and in what form.
What a Controller Actually Does
The controller title gets used loosely at startups, but in a well-run finance organization the role has a specific shape. A controller:
Owns the monthly close. The close is the process of finalizing the books each month: reconciling every bank and credit card account, recording accruals and deferrals, reviewing revenue recognition, posting payroll and equity entries, and producing financial statements. A good controller runs the close on a calendar, closes within five to ten business days, and delivers statements the rest of the company can rely on.
Owns accuracy. When the board asks whether the revenue number is right, the controller is the person whose job it is to say yes and mean it. That covers GAAP treatment of tricky items: deferred revenue, capitalized software, stock compensation expense, sales tax liabilities.
Builds and enforces internal controls. Who can approve a vendor payment, who can access the bank account, how expense reports get reviewed, how the company prevents duplicate payments and fraud. Controls sound bureaucratic until the first time they catch something.
Manages compliance. Sales tax registrations and filings, 1099s, state franchise taxes, payroll tax coordination, and supporting the CPA on income tax. At later stages, the controller is also the primary point of contact for financial statement audits.
Manages the accounting team. Bookkeepers, staff accountants, accounts payable and accounts receivable clerks. The controller hires them, reviews their work, and is accountable for their output.
Notice what is not on this list: fundraising, forecasting, pricing strategy, board narrative, capital allocation. A controller can be excellent, and most good ones are, without ever building a financial model or sitting across from an investor.
What a CFO Actually Does
A CFO operates one level up, as a member of the executive team. We covered the role in depth in our guide to the fractional CFO model from seed to exit, but the summary:
Strategy and capital allocation. Should we hire two more engineers or extend runway? Raise now or at the next milestone? Enter the new market or double down? These are financial decisions dressed up as operating decisions, and the CFO is the executive who frames them with numbers.
Fundraising. Structuring the round, building the model that supports the valuation narrative, preparing the data room, managing diligence, and negotiating terms alongside the CEO. For a venture-backed company, this is often the single highest-leverage thing a CFO does.
Forecasting and financial modeling. A bottoms-up model of revenue, expenses, hiring, and cash across scenarios. The model is not a spreadsheet artifact; it is the operating plan expressed in numbers, and it drives every hiring and spending decision.
Cash and runway management. Weekly cash visibility, a 13-week cash flow forecast, and early warning when the plan and reality diverge. Startups rarely die from bad accounting. They die from running out of cash, usually with less warning than the founder expected.
Board and investor relations. The monthly investor update, the quarterly board package, and the metrics narrative that connects what happened to what the company will do next. Investors calibrate their confidence in a company partly on the quality of this communication.
A CFO relies on accurate books to do any of this well. But producing those books is not the CFO's job, which is exactly why the two roles get confused: each one is incomplete without the other function existing somewhere.
The Comparison at a Glance
| Dimension | Controller | CFO |
|---|---|---|
| Orientation | Backward-looking, internal | Forward-looking, external |
| Core question | Are the numbers right? | What should we do next? |
| Owns | Close, controls, compliance, accounting team | Strategy, fundraising, forecasting, board |
| Produces | Accurate financial statements | Models, forecasts, board packages, the fundraise |
| Reports to | CFO (or CEO if no CFO) | CEO |
| Typical background | CPA, audit or accounting management | Investment banking, FP&A leadership, prior CFO or VP Finance |
| Full-time market comp | Roughly $150K to $250K | Roughly $200K to $400K, up to $500K+ at growth stage |
Compensation figures are typical market ranges for venture-backed companies, and they vary with geography, stage, and equity mix. The pattern holds everywhere: a CFO costs meaningfully more than a controller, and both cost more than most early-stage startups should spend on a full-time seat.
The Sequencing Question Generalist Articles Get Wrong
Search "CFO vs controller" and most articles, written for small businesses in general, land on the same advice: hire a controller first, because you need accurate books before you need strategy, and add a CFO later when you are "big enough."
For a bootstrapped services firm or a family-owned distributor, that sequencing is often right. Their financial complexity is mostly transactional, their stakeholders are internal, and nobody is asking for a Series A deck.
For a venture-backed startup, the sequencing is usually backward, for one structural reason: the CFO-shaped problems arrive before the controller-shaped problems do.
Consider what a venture-backed company actually faces in its first three years:
- A pre-seed or seed raise, which requires a financial model, a use-of-funds narrative, and defensible unit economics
- Runway decisions with real consequences, because the company is burning cash by design
- Investor updates and, after the first priced round, board reporting
- Pricing, hiring pace, and market decisions that each move runway by months
- A Series A process where the quality of the financial narrative directly affects valuation
Every one of those is CFO work. Meanwhile, the accounting reality at the same stage is comparatively simple: one entity, one currency, a few hundred transactions a month, straightforward payroll, and revenue that is either zero or early enough that recognition policy can be set once and followed. That workload does not need a $200K controller. It needs a competent bookkeeper, a CPA for taxes, and someone senior enough to review the output.
The generalist advice fails startups because it assumes financial complexity grows transactionally, from volume. At venture-backed companies, complexity arrives strategically first, from fundraising and burn, and only later transactionally.
The Right Order for Most Venture-Backed Startups
Here is the sequencing we recommend, and the one we see work across our client base:
Step 1: Bookkeeper (from day one)
Before either title matters, you need clean books. A bookkeeper handles transaction categorization, bank and credit card reconciliations, accounts payable and receivable, and monthly financial statements. Pair them with a CPA for income tax filings and you have the transactional layer covered for a small fraction of a controller's cost.
Skipping this step is the actual first mistake. No CFO or controller can work from books that do not exist, and cleaning up two years of neglected QuickBooks is more expensive than maintaining it would have been.
Step 2: Fractional CFO (typically seed, sometimes pre-seed)
When the CFO-shaped problems show up, and for most startups that is at or before the seed round, the answer is not a $200,000 to $400,000 full-time hire. It is a fractional CFO: the same executive-level capability, engaged 15 to 40 hours per month depending on stage, at a fraction of full-time cost.
At this stage the fractional CFO builds the model, runs the fundraise support, manages runway, produces board reporting, and, critically for this article, supervises the accounting function. They set the close calendar, define accounting policies, review the bookkeeper's monthly output, and catch the classification and revenue recognition issues a founder would never spot. That supervision is the controller-level judgment layer, delivered in a few hours a month rather than a full-time salary.
If you are unsure whether you have hit this point, our guide on when to hire a fractional CFO walks through the specific triggers, and our breakdown of fractional CFO cost covers what the engagement models look like in dollars.
Step 3: Full-time controller (typically around Series B)
There is a point where the accounting workload genuinely demands a dedicated owner. For most venture-backed startups that arrives around Series B, when several of these are true:
- Transaction volume has grown to thousands per month across multiple systems
- You have multiple entities, and possibly multiple currencies, to consolidate
- Revenue recognition requires real judgment: multi-element contracts, usage-based pricing, ASC 606 complexity
- Your investors or your debt provider require a financial statement audit
- Headcount has passed 50 to 75 and payroll, equity, and benefits accounting has real depth
- The close is consuming so much of the outsourced team's month that reporting timelines slip
At that point, hire the controller. They take over the close and the accounting team, and they free the CFO function, whether fractional or by then full-time, to stay at the strategic altitude where it earns its cost.
Step 4: Full-time CFO (when the work is full-time)
Many companies convert the fractional CFO relationship into a full-time hire somewhere between Series B and Series C, when board cadence, financing complexity, and team size make finance leadership a daily job. Others keep the fractional model longer. The controller hire does not force this decision either way.
What Each Path Costs
Using typical market ranges for venture-backed companies:
| Option | Annual cost | What it covers |
|---|---|---|
| Bookkeeper (outsourced) | Low four figures to low five figures | Transactions, reconciliations, monthly statements |
| CPA (tax engagement) | Low four figures to low five figures | Income tax filings, tax planning |
| Fractional CFO | Mid four figures to low six figures, scaling with hours | Strategy, fundraising, forecasting, board, accounting oversight |
| Full-time controller | ~$150K to $250K total comp | Close, controls, compliance, accounting team management |
| Full-time CFO | ~$200K to $400K, $500K+ at growth stage | Everything in the CFO section, full-time |
The comparison that matters is not controller versus CFO. It is the bundled early-stage stack (bookkeeper plus CPA plus fractional CFO) versus either full-time hire. The bundle typically costs less than half of a full-time controller alone, and it covers both sides of the function rather than one.
A Decision Framework by Symptom
Titles are abstract. Symptoms are not. Here is how to read what your company is actually telling you:
Your books are late, messy, or nonexistent. You do not have a controller problem or a CFO problem yet. You have a bookkeeping problem. Fix the transactional layer first; everything else depends on it.
Books are current, but you cannot answer investor questions. You know what you spent, but not your burn multiple, your runway under the new hiring plan, or what your CAC payback implies about the next raise. That is a CFO gap. A controller will not close it, because it is a forecasting and narrative problem, not an accuracy problem.
You are three months from starting a raise. CFO gap, urgently. The model, the metrics story, and the data room are CFO deliverables, and building them well takes months, not weeks.
Month-end close takes three weeks and the numbers keep getting restated. If your volume is genuinely high, multiple entities, complex revenue, audit on the horizon, that is a controller gap. If your volume is modest and the close is late anyway, it is more likely a bookkeeping quality problem plus an oversight gap that fractional CFO review solves.
Your auditors, or your Series B lead, are asking about controls. Controller territory. Segregation of duties, documented processes, and audit support are what the role exists for.
You have both problems at once. Most startups do, in some proportion. The answer is still sequencing: cover the accounting side with bookkeeper plus CPA plus oversight, and put senior dollars toward the strategic side, because that is where the leverage is at venture speed.
How a Fractional CFO Plus Accounting Team Covers Both Sides
The reason the startup sequencing works is that the two functions do not need to be two full-time people early on. They need to be two layers:
The execution layer is transactional: categorize, reconcile, file, pay. A good outsourced bookkeeping team and a CPA handle this reliably at early-stage volume.
The judgment layer is supervisory and strategic: set policies, review output, catch errors, then use the resulting numbers to forecast, fundraise, and report. A fractional CFO provides this in the same engagement that covers strategy, so the marginal cost of controller-level oversight is close to zero.
This is exactly how we structure StartupCFO: a bundled team of bookkeeper, CPA, and fractional CFO, with plans starting at $179 per month, so the execution layer and the judgment layer arrive together rather than as separate hires you have to sequence and manage yourself. The bookkeeper keeps the books current, the CPA keeps you compliant, and the fractional CFO reviews the output and turns it into forecasts, board reporting, and fundraise support. When your complexity eventually justifies a full-time controller, you hire into a clean, well-documented function instead of a cleanup project.
The Bottom Line
A controller and a CFO solve different problems. The controller makes your numbers trustworthy. The CFO makes them useful. Generalist hiring advice says controller first, and for a typical small business that is fine. For a venture-backed startup, the strategic problems arrive years before the transactional ones, so the order that works is bookkeeper first, fractional CFO second, and a full-time controller around Series B when transaction complexity genuinely demands one.
If you are trying to figure out which gap you have, the symptom test is simple: books late means bookkeeping, investor questions unanswerable means CFO. If it is the second one, or both, book a free consultation and we will walk through what the right finance stack looks like for your stage.
