The Best KPIs for General Contractors in 2027

A $40 million build that looked profitable until closeout

Picture a commercial general contractor with roughly $180 million in annual revenue, running eleven active jobs. The flagship project is a $40 million ground-up medical office building. Nine months in, the internal dashboard is green: the WIP schedule shows 15.2% gross margin, billings are current, and the project executive is confident enough to forecast the fee into the annual bonus pool.

Then closeout arrives and the number collapses to 8.1%. Nothing dramatic happened. Four things happened quietly and simultaneously, and none of them were visible because the company was measuring the wrong things on the wrong cadence.

First, the project team had been performing change-order work on verbal owner directives. Roughly $1.6 million of scope was executed against pending change orders that were never converted to signed COs. The cost hit the job cost ledger in real time; the revenue sat in a "pending" column that the margin calculation was happily including. When the owner's construction manager rejected $900,000 of it at closeout as "within the base scope of the contract documents," that margin was never real. It was an estimate the company had been reporting as fact for three quarters.

Second, the schedule slipped. Not catastrophically — about five weeks against a 62-week baseline, which is roughly an 8% variance. But the company updated its schedule monthly, so the slip surfaced in month four when the float was already consumed. Recovery meant premium-time drywall and MEP crews, an extra five weeks of general conditions burn at roughly $95,000 per month, and an accelerated closeout that generated its own punch-list cost.

Third, a mechanical subcontractor holding an $4.2 million scope went sideways. Nobody was monitoring sub financial health, so the first signal was a supplier lien filed against the project. Rebuying and remobilizing that scope mid-project cost multiples of the original contract — the industry rule of thumb runs two to four times the remaining subcontract value once you account for remobilization, schedule impact, and the premium a replacement sub charges to walk into someone else's half-finished work.

Fourth, retainage aged. The owner held 10% through substantial completion and then sat on $4 million for 140 days while the punch list and closeout documentation dragged. Meanwhile the general contractor had already paid most of its subs. The company was, functionally, lending the owner four million dollars at zero interest while drawing on its own credit line at prime plus two.

Every one of those four failures maps directly to a metric that a disciplined general contractor tracks. None of them are exotic. The reason they went undetected is that the company measured revenue and backlog — the two numbers everyone asks about at industry dinners — and treated everything else as accounting exhaust. The KPIs below exist specifically to catch these four failure patterns before they compound.

How the metrics actually chain together

The mistake most general contractors make is treating KPIs as a scorecard — a list of numbers reviewed once a quarter, each independent of the others. They are not independent. They form a directed chain where a degradation in one upstream metric propagates predictably downstream, usually with a two-to-four-month lag that makes the causal link hard to see if you are only looking at a monthly report.

The chain starts with schedule. Project schedule variance is derived from earned-value methodology: you compare the value of work actually put in place against the value your baseline schedule said should be in place at this date. When earned value falls behind planned value, three things follow automatically. General conditions — supervision, trailers, temporary power, safety staff, cleanup — continue to burn at a fixed monthly rate against a fixed contract amount, so every week of slip is pure margin erosion. Extended overhead exposure grows. And if the contract carries liquidated damages, you are now accruing a contingent liability that your WIP schedule may or may not be reserving against.

Schedule slip also drives change-order volume in both directions. Slipped jobs generate legitimate time-impact claims against the owner, and they generate scope disputes as trades stack on top of each other in compressed sequences. So change-order revenue percentage is downstream of schedule variance, not independent of it.

Change orders and subcontractor cost ratio jointly determine gross margin, because in a typical commercial general contractor's cost structure, subcontractors are the dominant line and change orders are the swing factor. Gross margin then determines cash generation, which combines with billing discipline and retainage practice to set the cash conversion cycle. Cash conversion cycle drives working capital days. And working capital days, along with safety performance, drives bonding capacity — because sureties underwrite on working capital, net worth, and experience-modification rate before they underwrite on anything else.

Bonding capacity closes the loop back to backlog, because a general contractor that cannot get bonded cannot bid the work that fills the pipeline. This is the part operators miss: a safety incident in Q1 raises the EMR at the next policy period, which shrinks the surety's single-job and aggregate limits, which excludes the company from bidding the exact large projects that carry the best margins. A safety metric becomes a revenue metric through the bonding channel.

Reading the chain backward is how you diagnose. If gross margin is compressing, do not start by interrogating the estimating department. Walk upstream: is the subcontractor cost ratio drifting? Is change-order revenue running hot or cold? Is schedule variance negative across multiple jobs? The margin number is a symptom that shows up last. The schedule and sub-cost numbers are the leading indicators that show up first, which is precisely why they need a weekly cadence while margin can survive on a monthly one.

Real numbers, ranges, and where the benchmarks come from

Every number below should be read as a band, not a target, and every band shifts by project type. A public infrastructure general contractor and a private tenant-improvement general contractor are different businesses that happen to share a license classification.

Gross margin. Formula: (revenue minus direct job cost) divided by revenue, where direct job cost includes subcontracts, materials, direct labor and burden, equipment, and bond premium. Residential general contractors typically run in the high teens to mid-twenties. Commercial general contractors run roughly 10% to 20% depending on delivery method and self-perform mix. Public and heavy-civil work runs lower, commonly in the high single digits to low teens, because the bidding is open, the scope is well-defined, and the competitive field is deep. Track it at job level monthly off the WIP schedule and company-wide quarterly. The single most useful refinement is to report two versions side by side: margin including pending change orders, and margin on signed contract value only. The gap between them is your exposure.

Project schedule variance. Formula: (earned value minus planned value) divided by planned value. Well-run commercial general contractors hold within roughly plus or minus 5%. Median performance across the industry is meaningfully wider. A variance worse than negative 15% on a job with liquidated damages should trigger an immediate reforecast and a reserve conversation with the CFO, not a note in the monthly package. Compute it weekly per active project.

Change-order revenue percentage. Formula: approved change-order revenue divided by original contract value. New ground-up commercial construction with complete drawings sits comfortably in the 5% to 10% band. Renovation, adaptive reuse, and infrastructure routinely run 10% to 25% because unknown existing conditions are structural to the work, not a failure of estimating. What matters is the direction and the conversion rate, not the absolute level: track pending change orders separately from approved ones and measure the PCO-to-CO conversion percentage and the average days from PCO issuance to owner signature. A conversion rate below 70% or an average age above 45 days is a broken process, regardless of what the headline percentage says.

Subcontractor cost ratio. Formula: total subcontract cost divided by total job cost. Pure construction-management-at-risk firms that subcontract nearly everything run high, often 70% or more. Typical commercial general contractors run roughly 50% to 65%. General contractors with meaningful self-perform capability — concrete, steel erection, carpentry, sometimes MEP — pull that down into the 30% to 50% range and capture the trade margin themselves. The signal is not the level but the quarter-over-quarter drift. A move of five points in a single quarter without a corresponding shift in project mix usually means scope is migrating out of self-perform for reasons nobody documented, or a sub is being asked to absorb scope that will come back as a claim.

Backlog in months. Formula: unearned signed contract revenue divided by trailing-twelve-month revenue divided by twelve. ABC publishes its Construction Backlog Indicator monthly, and it has hovered in the eight-to-nine-month range in recent readings. A healthy floor for most commercial general contractors is around six months; below four months, the bid department is behind and the field will be idle within two quarters. The critical discipline: only signed contracts count. Letters of intent, verbal awards, and "we're on the shortlist" are pipeline, not backlog, and mixing them corrupts the number your surety and your bank are reading.

Working capital days. Formula: days sales outstanding plus days inventory outstanding, minus days payables outstanding. Sixty to ninety days is a workable range for commercial work. Beyond 120 days, the company is financing its owners and will need a revolver draw or an equity injection to keep growing. Retainage deserves its own sub-metric — track retainage receivable as a percentage of total AR and track aged retainage over 90 days past substantial completion as a standalone dollar figure reviewed at every monthly close.

Safety TRIR. Formula: recordable incidents times 200,000, divided by total hours worked. The construction-sector average per BLS data sits meaningfully above 2.0. Strong commercial general contractors operate well below 1.0. The reason to care beyond the ethical one is mechanical: EMR above 1.0 disqualifies bids on a large share of Class-A commercial and federal work, and it raises workers' compensation premium on every hour worked. Pair the lagging TRIR with leading indicators — near-miss reports per 100,000 hours, job hazard analysis completion rate, and safety observation counts — because near-miss volume moves months before TRIR does.

Cash conversion cycle. Formula: days inventory outstanding plus days sales outstanding, minus days payables outstanding, computed at project level rather than company level. Private commercial work should target under 45 days. Public-works general contractors routinely run 90 to 120 days because government payment cycles are slow and prompt-payment statutes have long fuses. The lever is the schedule of values: a front-loaded, defensible SOV that recognizes mobilization, submittals, and long-lead procurement legitimately moves the whole curve left.

Revenue per field employee. Formula: annual revenue divided by field headcount only — foremen, journeymen, apprentices, laborers. Exclude project managers, estimators, and office staff or the metric loses all operational meaning. Self-perform civil contractors run lower because they carry more bodies against the same revenue. Construction-management-heavy firms run dramatically higher because the labor sits on subcontractor payrolls. The useful comparison is against your own prior-year figure and against firms with a similar self-perform mix, never against an industry average that blends both models. With ABC forecasting the industry needs hundreds of thousands of net new workers to meet demand, this metric is the cleanest available proxy for whether your labor strategy is holding.

Trade-offs, and the KPIs you deliberately choose not to chase

Every one of these metrics can be optimized to the point of self-harm, and the KPIs actively fight each other. A general contractor that does not understand the trade-off structure will improve one number and quietly break two others.

Margin versus backlog. The fastest way to raise gross margin is to bid selectively and refuse thin work. The fastest way to raise backlog is to bid aggressively and buy market share. You cannot maximize both at once with a fixed estimating capacity. The correct posture depends on where you sit: if backlog is above nine months, tighten bid margins upward and let the low-margin work go to competitors. If backlog is under five months, accept thinner work to keep crews intact, because the cost of losing a superintendent and two foremen to a competitor exceeds the margin you saved by declining a break-even job.

Subcontractor cost ratio versus risk transfer. Pushing the sub ratio down through self-perform captures trade margin, but it converts variable cost into fixed cost. A general contractor self-performing concrete owns the crew, the forms, the pump, and the payroll whether or not the next job has concrete in it. Sub ratio is a risk dial, not a performance score: high ratio means low fixed cost and low margin capture; low ratio means the opposite. Firms that self-perform successfully typically do it in one or two trades where they have genuine depth, not opportunistically across the board.

Cash conversion cycle versus subcontractor relationships. The easy way to compress CCC is to stretch payables — pay subs slower while collecting from owners at the same speed. It works for exactly as long as your subs have alternatives. In a market where good subs are the scarce resource, a general contractor known for slow payment gets the B-team crew, the padded bid, and the sub who walks when a better job appears. The durable version of CCC improvement is faster billing and faster collection, not slower paying.

Change-order percentage versus owner relationship. A general contractor can drive change-order revenue up by aggressively pursuing every ambiguity in the documents. That is legitimate — you are entitled to be paid for work outside the base scope. But an owner or architect who experiences your firm as claim-driven will not shortlist you next time. The firms that get repeat negotiated work tend to run moderate change-order percentages with very high conversion rates: they ask for less, but almost everything they ask for gets signed quickly.

Schedule variance versus safety and quality. Recovering a slipped schedule means acceleration: more crews in the same space, longer shifts, weekend work. Every one of those raises incident probability and rework rate. A general contractor that treats PSV as the single dominant metric will systematically trade TRIR for schedule and pay for it two policy periods later in EMR and premium.

There is also a meta trade-off about the number of metrics themselves. Nine KPIs is roughly the ceiling for a general contractor of any size. Firms that build forty-metric dashboards get worse, not better, because nobody owns forty metrics and the signal drowns. Assign exactly one accountable owner per metric — the project manager owns schedule variance and change-order conversion, the project executive owns job gross margin, the CFO owns working capital and cash conversion, the safety director owns TRIR and leading indicators — and review the owned number, not the dashboard.

The pitfalls that reliably destroy the numbers

Reporting margin on unsigned change orders. This is the most common and most expensive error in general contracting. The fix is structural, not behavioral: the WIP schedule should have two columns, signed contract value and pending change-order value, and the reported margin should be computed on signed value only, with the pending column shown as separate upside. Any project executive who wants credit for pending work has to explain the conversion probability out loud.

Monthly schedule updates. A monthly cadence guarantees you learn about a slip four to six weeks after it started, by which point recovery costs several times what it would have cost in week one. Move the schedule update and cost-to-complete review to weekly, keep it to thirty minutes, and require the superintendent — not the project engineer — to present the two-week look-ahead against the baseline.

Counting letters of intent in backlog. Inflating backlog feels harmless internally and is dangerous externally. Sureties and banks make credit decisions on that number, and a backlog that drops sharply because unsigned work evaporated reads as a business in decline, not a definitional cleanup. Publish one backlog number, signed and notice-to-proceed only, and keep pipeline in a separate report.

No subcontractor financial health monitoring. Most general contractors qualify subs once at prequalification and never look again. Build a lightweight recurring check: lien filings against the sub's other projects, payment-velocity signals from your own AP data, changes in bonding, crew-size changes on your site, and requests for out-of-sequence payment. Any two of those firing simultaneously warrants a conversation before it becomes a default.

Aging retainage. Retainage is money you have already earned. Assign one person accountability for retainage collection with a hard cadence: closeout documentation packages assembled before substantial completion, not after; retainage over 90 days escalated to the project executive; retainage over 150 days escalated to counsel. The cost of the awkward phone call is always lower than the cost of the revolver draw.

Mixing field and office headcount. Revenue per employee computed across the whole company tells you nothing actionable, because a shift in self-perform mix moves it without any change in productivity. Keep the field-only denominator, publish it quarterly, and compare it only against your own prior periods.

Benchmarking against the wrong peer set. CFMA benchmark data is genuinely useful, but only when filtered to your own revenue band, delivery method, and geography. A $50 million tenant-improvement general contractor comparing itself to the ENR-ranked national firms will draw exactly the wrong conclusions about sub ratio, overhead, and revenue per employee.

A workable rollout is ninety days. In the first thirty, rebuild the WIP schedule from the general ledger, standardize cost codes to CSI MasterFormat so job-to-job comparison is meaningful, and assign one named owner per metric. In the next thirty, move schedule variance and cost-to-complete to weekly, implement a digital change-order log with PCO-to-CO conversion tracking, and stand up subcontractor financial monitoring. In the final thirty, build the quarterly benchmark comparison, wire bonding-capacity utilization to the backlog forecast, and run the first board read-out with all nine numbers against peer median.

Related questions

How many KPIs should a general contractor actually track?

Nine or fewer at the company level, with one named accountable owner per metric. Larger dashboards fail because ownership dilutes and signal drowns. Project teams may track additional operational detail, but the executive review set should stay small enough that every number gets discussed.

What is the difference between gross margin and fee on a construction job?

Gross margin is revenue minus all direct job cost, divided by revenue. Fee is the contractually negotiated general contractor markup, typically a defined percentage in a cost-plus or GMP contract. On GMP work, realized gross margin can exceed or fall below the stated fee depending on buyout savings and shared-savings terms.

Why does safety performance affect a contractor's revenue?

Through bonding and prequalification. A rising experience-modification rate shrinks surety capacity and disqualifies bids on much Class-A commercial and federal work. Safety is a lagging cost metric and a leading revenue-eligibility metric simultaneously.

Should backlog include work awarded but not yet under contract?

No. Count only signed contracts with notice to proceed. Awarded-but-unsigned work belongs in pipeline reporting. Mixing the two corrupts the number your surety, bank, and board rely on, and produces alarming apparent declines when unsigned work does not convert.

How often should the WIP schedule be reviewed?

Monthly at minimum, at every close, with project managers presenting cost-to-complete rather than accounting deriving it. Weekly cost-to-complete reviews on active jobs catch variance while recovery is still cheap; the formal WIP cut can stay monthly.

FAQ

What gross margin should a commercial general contractor target?

Commercial general contractors generally operate in a 10% to 20% band depending on delivery method, self-perform mix, and market. Firms with meaningful self-perform capability sit toward the upper end because they capture trade margin. Public and heavy-civil work runs lower. The more important discipline than hitting a specific number is reporting margin on signed contract value only, so pending change orders never inflate the figure.

When is change-order revenue percentage a warning sign rather than normal business?

On new ground-up construction with complete drawings, a figure well above 10% suggests scope definition or estimating gaps. On renovation and infrastructure work, 10% to 25% is structurally normal because unknown existing conditions are inherent. Watch the PCO-to-CO conversion rate and the average days to owner signature more closely than the headline percentage.

What backlog level indicates a healthy general contractor?

Roughly six to ten months of signed, notice-to-proceed work suits most commercial general contractors. ABC's Construction Backlog Indicator has recently run in the eight-to-nine-month range industry-wide. Below four months, bidding has fallen behind and field crews will be idle within two quarters. Above twelve months, verify you have the labor and supervision to actually deliver it.

Why should subcontractor cost ratio be watched quarter over quarter?

The absolute level mostly reflects your business model — construction-management-heavy firms run high, self-perform firms run low. The signal is drift. A five-point swing in one quarter without a corresponding change in project mix usually means scope migrated out of self-perform undocumented, or a subcontractor absorbed scope that will return later as a claim.

How should retainage be managed to protect working capital?

Track retainage receivable separately from standard AR and report aged retainage over 90 days past substantial completion as its own dollar figure. Assemble closeout documentation before substantial completion rather than after, and escalate on a fixed schedule. Unrecovered retainage is a leading cause of general contractor liquidity crises.

What cadence should each metric be reviewed on?

Daily for safety leading indicators and field labor hours by cost code. Weekly for schedule variance, cost-to-complete, the change-order log, and subcontractor payment status. Monthly for job gross margin off the WIP schedule, subcontractor cost ratio, working capital days, and cash conversion cycle. Quarterly for company-wide margin, backlog months, TRIR, and revenue per field employee against benchmarks.

Sources

• https://www.cfma.org/ — Construction Financial Management Association, publisher of the Construction Financial Benchmarker
• https://www.abc.org/News-Media/News-Releases — Associated Builders and Contractors, monthly Construction Backlog Indicator releases
• https://www.agc.org/ — Associated General Contractors of America, annual workforce survey and industry data
• https://www.bls.gov/iif/ — U.S. Bureau of Labor Statistics, Injuries, Illnesses, and Fatalities program (TRIR baselines)
• https://www.osha.gov/recordkeeping — OSHA recordkeeping rules and recordable-incident definitions
• https://www.enr.com/toplists — Engineering News-Record Top 400 Contractors rankings
• https://www.aia.org/resources — American Institute of Architects contract documents and change-order guidance
• https://www.census.gov/construction/c30/c30index.html — U.S. Census Bureau Value of Construction Put in Place
• https://www.fasb.org/ — Financial Accounting Standards Board, ASC 606 revenue recognition guidance
• https://www.sec.gov/edgar/search/ — SEC EDGAR filings for public contractor backlog and working-capital disclosures

Related on PULSE

• [The Best KPIs for Roofing Contractors in 2027](/knowledge/ik0431)
• [The Best KPIs for HVAC Contractors in 2027](/knowledge/ik0429)
• [The Best KPIs for Electrical Contractors in 2027](/knowledge/ik0430)
• [The Best KPIs for Plumbing Contractors in 2027](/knowledge/ik0428)