Last Updated on May 28, 2026 by Team TBH
A development company doesn’t build anything. That isn’t a clever framing; it’s literally how the business works. The general contractor, the developer, and the construction manager at the top of any meaningful project produce almost no direct labor on the structure that gets built. What that company produces is coordination: across a temporary organization of specialty trade contractors, each operating under its own licensing regime, its own insurance profile, its own payroll structure, and its own workforce comp class codes. The wire pull is done by electrical subs. The pipe is run by mechanical and plumbing subs. The framing is done by carpenters who often work as separate entities. The HVAC is its own ecosystem. The development business is the connective tissue that makes those entities function as a single project, on a single schedule, against a single budget, under a single contract with the owner.
This is a much harder business problem than it looks from the outside. And it’s also a large one. BLS construction industry data puts the U.S. construction sector at roughly 801,000 establishments employing 8.3 million workers and generating about $3.1 trillion in annual revenue, roughly four percent of GDP. The structural detail that matters most for understanding how the business works, though, isn’t the sector total. It’s the fragmentation underneath: the top 50 specialty trade firms in the country represent only about seven percent of segment revenue. There is no equivalent of the big-tech consolidation pattern here. Every meaningful project assembles a new temporary org from a fragmented pool of small businesses, each of which functions as its own independently licensed, independently insured, independently capitalized operating unit.
What follows walks through what that coordination work actually consists of, and why it produces both the operational difficulty and the margin structure of a development business.
Why Coordination Is the Product
A development company’s competitive position is not built on doing the trade work better than its subs do. The trades are licensed precisely because the floor on competence is enforced by inspection; below a certain quality threshold, the work doesn’t pass and doesn’t get paid. The differentiation happens above the trade layer. It happens in how cleanly the GC can sequence three or four or fifteen specialty trades against a fixed schedule, how accurately it can predict and price the labor those trades will burn, and how reliably it can absorb the variance when one trade runs over and pushes the next one into overtime.
The coordination cost is not abstract. It shows up in scheduling software, in field supervision payroll, in change-order administration, in the office overhead of running a permit calendar and an inspection calendar across multiple jurisdictions, and increasingly in the software stack that runs underneath all of it. The shift toward construction payroll and job costing software illustrates what’s happening at the back-office layer: platforms purpose-built for the multi-trade, multi-rate, multi-site reality of construction labor are replacing generalist payroll tools that quietly assume every employee on the payroll runs the same wage profile under the same comp code at the same address. That assumption breaks the second a development company has crews from three subs working on the same site under three different prevailing wage classes.
The rest of the operational stack runs into the same problem in different shapes. Three of those shapes, namely licensing, insurance, and payroll, are the ones that most directly determine whether the coordination layer holds together or quietly leaks margin.
The Licensing Patchwork
Each trade in a development project operates under its own credentialing regime. Electricians run through state-administered licensure that typically requires four to five years of apprenticeship hours under a master electrician, followed by a journeyman exam and eventually a master license. Plumbers run through a structurally similar path under a different licensing body. HVAC contractors, in many states, carry their own mechanical license, sometimes split between installation and refrigerant handling under federal EPA Section 608 certification. General contractors carry their own license at the state level in roughly half of U.S. states, with most of the rest licensed at the city or county level. Each license has its own renewal cadence, its own continuing education requirements, and its own scope-of-work boundaries.
For a development company, the licensing patchwork creates a verification problem rather than a credentialing problem. The GC doesn’t perform the trades; it has to confirm that every sub it brings onto a project carries the right license for the work being done in the jurisdiction where the work is being done. A sub working out of license, because the master retired, the bond lapsed, the renewal got missed, or the state credential doesn’t cover the city in which the project sits, is a liability that flows up to the GC. Bid evaluation is partly a price exercise and partly a compliance audit, and the audit is only as good as the GC’s process for tracking sub credentials over time.
Permits add a parallel layer on top. Every electrical, plumbing, mechanical, and structural component of a project typically requires its own pulled permit, its own inspection schedule, and its own code compliance under the locally adopted version of the National Electrical Code, Uniform Plumbing Code, International Building Code, and so on. The permit calendar across a mid-sized project can run to dozens of inspections across multiple authorities. Missing one doesn’t just delay a trade; it stalls the entire sequence behind it.
The Insurance Patchwork
Where licensing is a verification problem, insurance is a stacking problem. Every entity on a construction site carries its own coverage profile, and that profile varies dramatically by trade. To take electrical work as the worked example, policy types electrical contractors carry reflects the dense coverage stack the trade actually requires: general liability for third-party injury and property damage, workers’ compensation (legally required in nearly every state for any electrical contractor with employees, and extended to sole proprietors in construction trades in several states because of the injury risk), commercial auto for service vehicles, tools and equipment coverage for inland marine exposure, professional liability for design errors, and surety bonds for licensing and project guarantees.
Plumbing carries a structurally similar but actuarially different stack: same categories, different premiums, different exclusions, different bond requirements. Roofing’s profile is even more divergent, with workers’ comp class codes among the highest-rated in the entire construction sector because of fall exposure. HVAC sits closer to electrical on the risk profile but adds refrigerant-handling exposure that has its own coverage implications. Concrete and excavation work pulls higher rates again for the heavy equipment and trench-collapse exposure.
The GC’s own coverage sits on top of all of that. A general contractor typically carries a commercial general liability policy, a builder’s risk policy specific to the project under construction, workers’ comp for its own employees, professional liability if it performs design-build work, and umbrella liability stacked above the primary coverages. The contract with the owner usually requires the GC to name the owner as an additional insured. The contract with each sub usually requires the sub to name the GC and the owner as additional insureds on the sub’s policy, and to provide certificates of insurance proving it. Tracking those certificates across an active project, with renewal dates that don’t line up, carriers that change mid-project, and limits that may or may not actually meet the contract requirements, is its own administrative function.
When this coordination fails, it fails expensively. A loss on a project where the responsible sub turns out to have let its policy lapse, or carried limits below the contract requirement, doesn’t just hit the sub. It rolls up the chain.
The Payroll Patchwork and Prevailing Wage
The most operationally complex of the three coordination layers is payroll. On a single project on a single day, a development business may have its own salaried supervisors on site, hourly employees doing punch-list and quality-control work, and crews from multiple subs running entirely separate payrolls under entirely different rate structures. The GC’s own labor goes through the GC’s payroll. The subs’ labor goes through theirs. But the cost data on every hour worked has to roll up to the GC for job costing to mean anything.
When the project touches federal money, the complexity compounds. Davis-Bacon Act prevailing wage rules require contractors and subs on federal construction contracts above $2,000 to pay laborers and mechanics no less than the locally prevailing wage and fringe benefit rate for their classification, and to submit certified weekly payroll reports (typically Form WH-347) documenting compliance. The same worker can carry different prevailing wage rates on different projects, or even at different points within a single project, depending on what classification of work they’re doing in a given hour. State-level “little Davis-Bacon” laws extend similar requirements to state-funded projects in many jurisdictions.
This is the level of operational detail at which generalist payroll software stops being able to keep up. A standard payroll tool processes a worker’s hours at a single rate, runs them through a tax engine, and writes a paycheck. A construction-vertical tool has to handle multiple rates per worker per pay period, allocate hours across multiple jobs and cost codes, assign the correct workers’ comp class code to each hour, generate certified payroll output in the format the agency requires, and reconcile it all into job-level cost reports that tell the GC whether the project is on budget. The same workers’ comp integration that pulls the right class code into payroll feeds back into the experience modifier that determines next year’s premium, which means a clean process at the payroll layer is a cost lever at the insurance layer.
Workforce Risk and the Cost of a Bad Year
The reason workers’ comp is such a structurally large line item in this sector is that the underlying risk is real. Construction and extraction occupations recorded 1,032 fatalities in 2024 in the BLS Census of Fatal Occupational Injuries, more than any other occupational group except transportation and material moving. Falls, slips, and trips alone accounted for 370 construction-sector deaths that year. Nonfatal injury rates in construction run materially above the cross-industry average, particularly in roofing, framing, and electrical work.
For a development company, those numbers are not abstract safety data. They are direct inputs into the cost of labor on every project. A claim against a sub’s workers’ comp policy raises that sub’s experience modifier, which raises the sub’s premiums, which raises the sub’s bid prices on the GC’s next project, which raises the GC’s costs. A clean multi-year safety record across the GC’s regular sub base is, in financial terms, a procurement advantage. A claim-heavy stretch on a particular trade is a price increase that the GC pays for years.
This is why safety programs at the GC level, including site-wide PPE enforcement, daily toolbox talks, incident documentation, and sub-level safety qualifications during bid evaluation, show up in the operating budget rather than the corporate-responsibility section of the annual report. They’re financial controls, not cultural initiatives.
Margins and What Separates the Operators
The economic profile of a development business reflects the structure of the coordination it has. Gross margins on managed construction work are thinner than most outside the industry assume, often in the single digits at the project level on hard-bid work and only modestly higher on negotiated and design-build work. The compensating factor is that the GC’s own direct labor and capital are minimal relative to project revenue. It’s a coordination business, not a labor business, which makes return on equity and return on overhead the relevant metrics, not gross margin per project.
What separates a development company that operates at the better end of those ratios from one that doesn’t is almost never sales volume. It’s whether the operational stack works. It’s whether the GC has a clean view of which subs are carrying current insurance and which aren’t. It’s whether labor hours across the GC’s own crews and across its subs’ crews are being tracked against job and cost code in something close to real time. It’s whether certified payroll on prevailing wage jobs is generated as a byproduct of normal payroll rather than reconstructed every Friday by an administrator. It’s whether the permit calendar and the inspection calendar across multiple jurisdictions are running on a system rather than in someone’s head.
When those things work, the GC has a cost view that lets it price the next project accurately and a compliance view that lets it sleep through audit season. When they don’t, both views are guesses, and the guesses tend to be wrong in the same direction: costs higher than estimated, compliance gaps that turn into back-wage liability, sub failures that turn into the GC’s failures.
The Software Stack Is Consolidating Under the Coordination Problem
The infrastructure most small and mid-sized development companies used to run their back office a decade ago, including spreadsheets, paper time cards, faxed certificates of insurance, and a local accountant doing weekly payroll filings, has been steadily replaced by category-specific software pitched directly at construction. This is part of a broader pattern in the SMB human capital management market, where cloud-native players have moved aggressively into vertical specialization. The competitive dynamics around Paychex’s competitors in HCM illustrate the shift: legacy service-bureau models built for generic small businesses are giving up ground to platforms that natively understand the operational shape of specific industries, construction among them.
The same procurement pattern shows up in adjacent categories. Digital insurance marketplaces have replaced the local broker for many small subs, letting them compare quotes from carriers in a single online application and update coverage as the business grows. That distribution model is analogous, at the SMB end of the market, to what Deel’s global payroll model did for cross-border employment: software replacing brokered service as the default procurement model for a category that used to require a phone call and a relationship.
For a development company, the practical upshot is that the operational stack underneath the coordination work is now buildable from a small number of integrated software products: a construction-vertical payroll and job costing platform, a digital insurance marketplace for the company’s own coverage, a tracking system for sub COIs, a project management and scheduling tool, a permit and inspection calendar, and an accounting system. Five years ago, this stack required either a much larger administrative headcount or a much larger ERP investment. The cost curve has come down, which is part of why mid-market GCs can now compete operationally with much larger firms.
The Business Underneath the Project
A development business doesn’t sell construction. It sells the ability to deliver construction on a schedule, on a budget, under a contract, across a temporary organization of specialty subs operating under entirely different rules. The licensing layer, the insurance layer, the payroll layer, and the workforce risk layer are not back-office concerns. They are the coordination work itself, expressed as compliance and accounting. A GC that runs them well has a procurement advantage, a cost advantage, and a margin advantage that competitors operating on spreadsheets simply cannot match for very long.
The trades will always be what gets built. The coordination is what gets sold. And the operational stack is what makes the coordination possible without bleeding margin every step of the way.
To read more content like this, explore The Brand Hopper
Subscribe to our newsletter
