What CFOs at Specialty Contractors Should Actually Measure Weekly

The monthly close is too late. Most CFOs at specialty contractors learn this the same way: by watching a project that looked fine in the third week of the month turn into a margin-fade write-down by the fifth. By then the labor was already burned, the change order window had already closed, and the only thing left to do was explain the variance.

Weekly metrics are not a replacement for the monthly close. They are an early warning system that protects the close from being a surprise. The trick is choosing the right ones. Specialty contractors do not have time for a dashboard with forty indicators. They need the few numbers that, when they move, mean something is happening on a project that the office should know about before the next billing cycle.

Below is a short list. None of these are exotic. What separates CFOs who run profitable specialty contracting businesses from CFOs who chase variances in arrears is rarely the sophistication of the metrics. It is the cadence and the data quality behind them.

Labor cost variance by cost code

The first metric on the list, and the one most likely to flag trouble before anything else does, is labor cost variance at the cost code level. Project-level variance is too coarse. A job can be 5 percent over on labor in aggregate while two cost codes are running 30 percent over and three are running under, and the executive summary will hide all of it.

Labor is the largest controllable cost on most specialty contracting projects. The Construction Financial Management Association, citing FMI research, notes that labor costs typically account for 15 to 30 percent of total project costs on heavy and civil work, and the share runs higher on most trade contractor jobs. CFMA also points out that cost-to-complete forecasting, which depends on accurate weekly labor data, is one of the markers that distinguishes higher-margin contractors from lower-margin ones.

Weekly variance at the cost code level should be the first report a CFO opens on Monday morning. Anything more than a 5 percent deviation on an active cost code is a conversation. Anything more than 10 percent is an action item.

Hours-to-budget by project, weekly

Closely related, but worth tracking separately, is hours-to-budget. This is the simpler version of the same question: are the hours being burned tracking against what was estimated? It is the leading indicator that catches problems before they show up in dollars.

The reason to keep it separate is that wage inflation, classification mix, and burden rates can all distort the dollar view. Hours are cleaner. If a crew estimated at 600 hours has burned 450 by the end of week three on a four-week scope, that is a problem regardless of what the dollar variance looks like.

This metric depends on knowing, with confidence, who was on the job and for how long. That is harder than it should be. The 2023 FMI Labor Productivity Study, which surveyed more than 250 senior leaders from labor-intensive self-performing contractors, found that 60 percent of respondents believe 11 percent or more of field labor costs are wasted or unproductive, with industry-wide losses estimated at $30 billion to $40 billion annually. The study also found that 79 percent of contractors believe they could improve labor productivity by 6 percent or more with better management practices. Most of that improvement starts with knowing where the hours actually go.

T&M billing readiness

For specialty contractors doing time-and-materials work, weekly T&M billing readiness is non-negotiable. The question is simple: of the T&M hours worked this week, what percentage are documented well enough to bill without dispute?

The reason this metric matters weekly, not monthly, is that the documentation gets harder to reconstruct as time passes. A foreman who can confirm a crew’s presence on a particular morning two days later may not be able to do it two weeks later. By the time a billing dispute reaches the GC’s project manager, the contractor’s leverage has eroded.

The CFOs who track this weekly tend to do so because they have been on the wrong end of a T&M dispute. Once is enough. Accurate weekly tracking of construction cost tracking data, including verified labor presence tied to specific cost codes, is what turns a billing conversation into a routine approval. When the data is clean, the GC has nothing to dispute. When it is not, the contractor absorbs the difference.

Days sales outstanding, by project

Aggregate DSO is a useful number for the bank. Project-level DSO is a useful number for the CFO.

Construction is, by the nature of how projects are billed and paid, a slow-paying industry. Reporting on the Hackett Group’s working capital research, CFO magazine noted that engineering and construction had the longest DSO of any tracked industry sector, and that overall DSO across the largest U.S. companies was 40.6 days in 2021. Specialty contractors sitting below GCs and owners in the payment chain typically run higher than that, sometimes considerably so, because of progress billing cycles, retainage, and multi-tier approval chains.

Aggregate DSO will not tell a CFO which project is dragging. Project-level DSO will. A weekly review that flags any project where DSO has crept past the firm’s threshold gives the CFO time to call the GC, escalate the pay app, or adjust cash forecasting before payroll becomes a question. The metric is not about collections pressure. It is about visibility. A project trending from 45 days to 65 days in two weeks is a different conversation than a project that has been at 65 days all year.

Headcount versus plan

Weekly headcount versus the staffing plan is a metric that gets dismissed because it sounds operational. It is operational. It also drives every other number on this list.

A project that was bid at a six-person crew and is running with eight will burn the labor budget early and produce a margin shortfall. A project bid at six and running with four will fall behind schedule and create overtime exposure later. Both situations show up in the labor variance report eventually. They show up in the headcount-versus-plan report immediately.

The CFO’s job is not to micromanage staffing. It is to ensure that when staffing diverges from the plan, the project manager can explain why, and that the explanation reaches the office before the variance does. A weekly check is enough.

Productivity unit ratios

For specialty contractors with measurable output, weekly productivity ratios are the most direct view of whether work is happening at the rate the bid assumed. Electrical contractors measure feet of conduit installed per labor hour. Concrete contractors measure cubic yards poured per crew-day. Drywall contractors measure square feet hung per hour. Solar contractors measure panels installed per crew-week.

These ratios are only as good as the data feeding them. The hours number has to be accurate, the unit count has to be accurate, and the two have to be tied to the same cost code. CFOs who try to build productivity dashboards on top of unreliable time data end up with dashboards that nobody trusts.

The CFMA piece on labor productivity makes this point in the context of cost-to-complete forecasting: the discipline only works when the underlying data is reliable. The same is true of productivity ratios. Weekly review only adds value when the weekly data is clean.

A note on what not to add

The temptation, once a weekly cadence is established, is to add metrics. Cash position. Backlog. Bonding capacity. Working capital ratio. These are all legitimate measures, and they all matter. They are also better suited to monthly or quarterly review.

The metrics that belong on a weekly list are the ones where a one-week movement actually means something. Backlog moves slowly. Bonding capacity moves slowly. Labor cost variance, hours-to-budget, T&M billing readiness, project-level DSO, headcount versus plan, and productivity ratios all move fast enough that a week of inattention can change the financial picture of a project. That is what makes them weekly metrics rather than monthly ones.

A CFO who tracks all six of these every week, with reliable data underneath them, has the information needed to ask the right questions in the project review meeting before the variance shows up in the close. The questions are usually not “what happened” but “what is happening, and what does the team need to do this week to keep it from getting worse.” That shift, from explaining variances to preventing them, is the whole point.

The data quality question, briefly

None of this works if the underlying data is unreliable. The CFO who builds a weekly cadence on top of timesheets reconstructed Friday afternoon, cost code assignments made from memory, and unit counts pulled from foreman text messages will eventually conclude that the weekly metrics are not telling the truth.

The fix sits ahead of the dashboard, not inside it. Time data needs to be captured at the moment of work, not reconstructed afterward. Cost codes need to be assigned at the worker level, not allocated in the office. Unit counts need to come from the field, not from someone’s best guess at percent complete. When the data is clean, the metrics work. When it is not, the metrics produce false confidence, which is worse than no metrics at all.

CFOs who have built reliable weekly cadences tend to talk about the project the same way: most of the work was making the data trustworthy. The metrics themselves were the easy part.

All external links verified functional on April 27, 2026.