OUTSOURCE CONSTRUCTION ACCOUNTING VS IN-HOUSE. THE REAL COMPARISON.
The outsource vs in-house decision for construction accounting gets framed the wrong way, as a cost comparison between a monthly fee and a salary. The real comparison is between what each option corrects. A $70,000 bookkeeper records transactions accurately. A $110,000 controller closes the books monthly and produces financial statements. Neither one finds an overhead rate problem, runs AR collections, produces cost to complete reports, or flags a job heading for a loss before closeout. That's a different scope conversation, and it's the one that decides the answer.
Most owners benchmark in-house accounting by salary, and salary is roughly half of the true load. Employer payroll taxes run 7.65 percent, health insurance runs $8,000 to $14,000 a year, and a 401k match adds another 3 percent. Add software, training, amortized turnover, and 4 to 6 hours a week of your own supervision time at $150 to $250 an hour, and the in-house controller at a $5M commercial subcontractor costs $145K to $238K fully loaded. Most owners price that same hire at $82K. This page runs the honest comparison across every option, an internal bookkeeper, an in-house controller, a full time CFO, a generic fractional CFO, and an outsourced construction CFO.
WHAT EACH ONE DOES.
| Capability | Internal Bookkeeper | In-House Controller | Full Time CFO | Generic Fractional CFO | Outsourced Construction CFO |
|---|---|---|---|---|---|
| Loaded annual cost | $55,000 to $75,000 | $100,000 to $140,000 | $175,000 to $250,000 | $24,000 to $60,000 | Flat monthly fee by revenue band |
| Weekly transaction entry and bank recs | Yes | Yes | No | No | Yes, entered weekly |
| Monthly close and financial statements | No | Yes, by the 10th | No | Some reporting | Yes, by the 10th |
| Job cost structure built to your trade | No, codes into what exists | Keeps what exists current | No | No | Yes, aligned to your estimate template |
| Monthly WIP, overbilling and underbilling | No | Sometimes | Reviews it | No | Yes, every month |
| Cost to complete on every active job | No | No | Yes | No | Yes, by the 12th |
| 13 week cash flow forecast | No | No | Yes | Some advisory | Yes, cash problems visible 8 weeks out |
| AR collections cadence | Records the aging only | No | Yes | No | Yes, calls at 31, 46, and 60 days |
| Construction specific method | Varies by hire | Varies by hire | Varies by hire | No | Yes, built for the $1M to $12M subcontractor |
| What it leaves uncovered | Job costing, WIP, collections, and CFO advisory | CFO advisory, cash forecasting, and PM accountability | A bookkeeper and controller underneath, $280,000 to $400,000+ all in | Construction method, job costing setup, and trade overhead structure | Nothing, no scope gaps |
Loaded annual cost includes salary, employer payroll taxes at 7.65 percent, health insurance of $8,000 to $14,000 a year, and a 401k match of 3 percent. It doesn't include recruiting cost, training time, or turnover risk. Add those plus software of $3K to $10K, training of $4K, and 4 to 6 hours a week of owner supervision at $150 to $250 an hour, and the in-house controller at a $5M subcontractor runs $145K to $238K fully loaded against a salary benchmark of $82K.
WHEN NOTHING IS WRITTEN DOWN YET.
An internal bookkeeper who records transactions accurately and closes out the month is doing the job correctly. The trouble is that recording transactions isn't the same work as managing the financial outcomes of the business. A bookkeeper records $180,000 in AR sitting at 65 days with no follow up process, because following up was never part of the job. They record the cost overrun on a project heading for a $45,000 loss at closeout, because catching it at month two was never part of the job either.
So the bookkeeper function is necessary and it's not sufficient. If the books are current and reconciled and you still can't tell which jobs earn or which week you run short, the bookkeeper is doing their work and the missing function sits above them. Hiring a second one produces the same reports from somebody new, and a $60,000 bookkeeper who doesn't know construction is more expensive than they look.
WHEN YOU NEED THE CLOSE TO HOLD.
A controller who closes the books by the 10th, produces accurate financial statements, and keeps the job cost structure current is doing excellent work. The limit is that controllership looks backward, reconciling what happened rather than heading off what's about to. A controller doesn't run the 13 week cash forecast, work the banking relationship before you need it, catch a job loss at month two, or run the AR collections cadence. Those are CFO functions, and between $1M and $8M of revenue a full time CFO stacked on top of a full time controller rarely pencils out.
There's also a single point of failure. Average tenure for a construction controller at a $5M subcontractor runs 2.5 to 3.5 years, replacement cost averages $25K to $40K, and the search itself runs 3 to 6 months before the ramp even starts. Amortized, that turnover line is $8K to $13K a year. If you already have a controller you like, the question isn't whether to replace them. The question is whether the CFO layer exists above them, and if it doesn't, the controller can keep the books while the outsourced function covers the forecast, the cost to complete, the collections cadence, and the monthly decisions.
WHEN YOU CAN CARRY THE WHOLE SALARY.
A full time construction CFO runs $175,000 to $250,000 loaded and gives you full financial leadership. What that salary doesn't include is bookkeeping or controllership, so the hire doesn't stand alone. You still need a bookkeeper and a controller underneath, which puts the true cost of the arrangement at $280,000 to $400,000 or more.
For a commercial subcontractor between $1M and $12M that number isn't defensible against the revenue it supports. The work is real and the salary isn't, which is the entire reason fractional structures exist for this size of company. Above roughly $20M to $25M the comparison gets closer to even, and above $25M a hybrid of in-house transaction staff plus outsourced CFO advisory is often the right build.
WHEN THE FINANCE IS RIGHT AND THE TRADE ISN'T.
A generic fractional CFO runs $24,000 to $60,000 a year and brings genuine financial expertise. What's missing is that commercial subcontracting has specific financial failure modes a generalist has no reason to know. Pay when paid billing cycles. Stored materials billing for electrical switchgear. WIP distortion from overbilling relative to completion. Equipment cost basis calculation. Job cost code architecture by trade.
None of those are general finance concepts. They're construction operating behaviors, and an implementation that doesn't account for them produces a financial system that looks correct and doesn't correct the underlying problems. You get oversight and some reporting without the job costing implementation or the trade specific overhead structure that would make the reporting mean something.
WHEN YOU WANT NO SCOPE GAPS.
This is the option built to leave nothing uncovered. Receipts, invoices, and timecards entered weekly rather than monthly. A job cost structure built to your trade with cost codes aligned to your estimate template. Monthly close by the 10th. Cost to complete on every active project by the 12th, so a job loss is caught at month two and not at closeout. A monthly WIP schedule that points at overbilling and underbilling. A CEO Report with 10 rolling metrics. A 13 week cash flow forecast that makes a cash problem visible 8 weeks out. AR collections management with a weekly aging review and a call cadence at 31, 46, and 60 days. A monthly strategic meeting that ends in three action items with owners and deadlines. Banking and bonding relationship work, including clean monthly financials and the line of credit conversation before you need it.
The work is also spread across a team instead of concentrated in one hire. The lead CFO owns the relationship, a bookkeeper covers transactional work, and a controller level resource covers reconciliations and WIP, so a vacation or a resignation doesn't stop the close. Onboarding runs 60 days, books migrated back to the start of your last taxable year, job costing rebuilt, and the close cadence installed, and after that the owner involvement is roughly 5 hours a month.
WHERE WE COME OUT.
The question isn't whether an outsourced construction CFO costs more or less than an in-house hire. The question is what each option corrects. A bookkeeper doesn't correct billing lag, overhead underrecovery, WIP distortion, or a job heading for a loss at closeout. A controller reconciles all of it after the fact. A full time CFO fixes the decisions and still needs two people underneath to produce the numbers. That's the comparison, and for a commercial subcontractor between $1M and $12M it comes out in one direction.
There's a risk answer on top of the cost answer. An in-house controller is one person. When they take vacation the books slow down, when they get sick deadlines slip, and when they leave the company loses institutional knowledge that takes the next hire months to rebuild. We have onboarded clients who came in after their controller left with two weeks notice and four months of unreconciled books behind them, with no documentation. For a $5M to $10M subcontractor the risk adjusted comparison favors outsourcing before you even count the money.
The honest read on the other side: under $20M, outsourcing wins on both cost and scope. Between $20M and $25M the comparison is close to even. Above $25M a hybrid is often the right answer, with transaction volume covered internally at clerk cost and the judgment work covered by specialists. If your books are genuinely behind, start by fixing the record, because none of the rest works without it.
FREQUENTLY ASKED.
