How to Choose a Fractional CFO for Contractors

Your best estimator is you.

That is the sentence most contractors will not say out loud, and it is the one that decides whether hiring a fractional CFO works or wastes a year.

We sat in a call with a commercial contractor where the question on the table was simple: could anyone else in the building explain why a job came in where it did? The answer was no. The owner held the true cost picture, including the cushions built into the estimating templates, and nobody had been trained to read it. The reports existed. The judgment behind them lived in one head.

That is the gap a fractional CFO is supposed to close. Most of them cannot, because they have never seen how contracting money actually moves.

What Should a Construction Company Look for in a Fractional CFO?

Look for three things: someone who can reconcile your estimate to your actuals at the job level, someone who understands that your revenue is a judgment rather than an event, and someone who can move the cost knowledge out of the owner's head and into a system. Everything else is secondary, including credentials, firm size, and rate.

Those three matter because they are where contracting money is actually made and lost. A general CFO reads your income statement correctly and still cannot tell you whether it is true, because the number was manufactured upstream in a percent-complete calculation they do not know how to audit.

The test is not whether they can read your statements. It is whether they can tell you which of your jobs is lying to you, and why.

Why Does Construction Break a Generalist CFO?

Because in contracting, revenue is recognized against an estimate of how complete a job is, so reported profit depends entirely on whether that estimate and the costs behind it are accurate. In most industries revenue is an event: something shipped, something was invoiced, cash arrived on terms. In construction it is a continuous judgment that gets trued up later.

Between the work and the money sit percent-complete mechanics, retainage held back by contract, change orders that may or may not have been papered, and draw schedules controlled by someone else. Each one distorts reported profit in a direction a generalist framework does not flag.

The practical consequence is timing. A generalist will find these distortions. They will find them after they have already shaped a quarter's decisions, and you will fund that education.

What Does It Look Like When Nobody Else Can Read the Numbers?

It looks like a business with real reporting where every meaningful judgment still routes through one person. Three situations, from three different construction companies, show the same structural hole in three different places.

A commercial contractor client of ours had project managers tracking mandates, man-hours, and billing thresholds. Careful work, done consistently. What they did not track was revenue, cost of goods, or gross margin against the job. Detailed cost visibility sat in a job-cost system that one person actually read. Separately, the estimating tools carried safety nets that had been built in years earlier and never documented, which meant perceived margins read lower than actual margins, and estimators priced defensively as a result. The company had to use incentive compensation to push bidding back up. Nobody had connected the two facts.

A specialty construction client of ours had the same hole at the front of the process. Their work mixes large, unique, highly variable activities, and there was no repeatable estimating model. Every significant bid was built from scratch, which is defensible on a one-off and a serious problem when a customer wants a complete quote inside forty-eight hours including travel, insurance, and overhead. Speed and accuracy were competing, and the institutional knowledge that would have resolved the tension had never been captured in a checklist or a cost-category template.

A past construction client of ours had solved the reporting layer and still could not close the loop. Their work-in-progress reviews ran monthly as a standing discipline, with real cost pooling behind them. What kept the month open was timing: costs arriving in the wrong period, and a close calendar that had never been locked. The schedule was mature. The inputs feeding it were not.

Why Can't My Accountant or Controller Do This?

Because you are asking them to do a job their role is not built for. A controller is accountable for whether the record is accurate. An accountant is accountable for compliance and filing. Neither is accountable for whether the resulting picture is a sound basis for a bid, and neither has the standing to tell you your estimating template is wrong.

Here is the honest version. In-house genuinely is the right answer sometimes. If you run one entity, your books are current, your job mix is consistent, and your decisions are mostly operational rather than financial, a strong controller plus a good accountant is a complete answer and adding a CFO is overhead. The threshold is not revenue. It is whether the financial decisions in front of you are big enough and frequent enough that getting them wrong costs more than the seat.

The signals that you have crossed it are specific. You cannot name which of your last twenty jobs beat its bid. Your work-in-progress schedule has never tied cleanly to your financial statements. You are the only person who can explain a variance. One of those is tolerable. All three at once means the decision layer is empty.

What Questions Actually Separate One Firm From Another?

Five, and none of them is about rate.

Ask who will personally be in your numbers each week, their seniority, and how many other clients they carry. A low quote usually means either a narrow scope or a junior person with a senior name on the proposal.

Ask how they will reconcile your estimates to your actuals, job by job, and what they will do when the two disagree.

Ask what they do in the first sixty days to establish that your record is reliable, and what they will tell you if it is not.

Ask them to explain your cash cycle back to you before you explain it to them. If retainage, draw timing, and change-order papering do not come up unprompted, they are going to learn your business on your dime.

Ask when they would tell you not to hire them. A firm that cannot describe the case against itself has no framework for fit, only a pipeline.

What Changes When the Cost Knowledge Leaves Your Head?

The estimating template stops being folklore. The cushions get named, measured, and either kept deliberately or removed, which means your estimators start pricing against reality instead of against a number they have learned to distrust.

Job variance arrives with an explanation attached rather than as a figure someone has to interpret. Percent complete is computed against a cost base you can defend. The month closes on a date rather than when the last correction lands.

The compounding effect is on decisions. You stop bidding the next job high to recover a margin problem that was an accounting artifact, and you stop bidding it low on a margin that was never really there. And the business stops being one person's memory, which matters more than most owners want to admit until the day it does.