A framer in Houston named Darrell used to get work through a foreman he had known for nine years. They spoke on the phone. Darrell showed up Monday, worked through Friday, collected a check. When the job ended, the foreman called again. That was the system.
Now Darrell opens an app at 4:30 AM. He scrolls shifts the way his daughter scrolls restaurants on DoorDash. A 12-hour framing gig in Katy pays $39/hour. Another in Sugar Land pays $36 but starts an hour later. He taps one, waits for a confirmation ping, drives 40 minutes. He has never met the general contractor. The general contractor has never met him.
Darrell is not a real person. He is a composite, assembled from interviews with platform workers and staffing industry reports. But his morning routine is real. It belongs to thousands of construction workers across the Sun Belt who now find jobs the same way people find rideshares. An algorithm scores them, ranks them, and decides if they work tomorrow. The industry calls this innovation. It might also be the largest quiet restructuring of construction labor in a generation.
480,000 Workers, 384,000 Jobs, Zero Matches
Start with the paradox that makes the platforms possible.
According to U.S. Chamber of Commerce data, the construction industry averaged roughly 383,900 job openings per month in 2023. In the same period, approximately 480,300 unemployed workers with construction experience were available. On paper, that is a surplus of 96,400 people. Contractors should have been swimming in applicants.
They were not. The Associated General Contractors of America reported in February 2025 that 80 to 90 percent of contractors still could not fill open positions. ABC, the other major trade group, estimated the industry needs 439,000 additional workers in 2025 to meet demand, and projected that gap could hit 520,000 by late 2026 if nothing changes.
The problem is not headcount. It is geography and specialization. An unemployed roofer in rural Ohio cannot help a framing crew in Phoenix. A finish carpenter in Chicago is not interchangeable with a concrete finisher in Dallas. The labor exists. The matching does not.
That number is my calculation, not an industry figure, and it rests on assumptions I will lay out. Average construction hourly earnings hit $36.54 per hour in early 2025, an 18% premium over private-sector averages, per AGC data. At 160 hours per month, each additional matched worker unlocks $5,846 in monthly labor value. Close 20% of the mismatch, and you get about 76,780 new placements monthly, or roughly $449 million per month. Annualized: $5.4 billion in labor value that currently evaporates because the right worker and the right job never find each other.
That is the pitch. Fill the gap with software.
Who Is Filling the Gap
Instawork has onboarded more than 8 million workers across industries and claims to fill 97% of shifts within 24 hours, compared to seven to ten days through traditional staffing agencies. Their AI screens workers with conversational vetting, checks certifications in real time, and matches based on ratings, shift history, and peer feedback. A predictive system flags at-risk shifts and auto-fills backups. Geofenced time tracking prevents "time theft." The language on their website reads like a logistics optimization problem. Which, for the platform, it is.
Workrise, backed by Andreessen Horowitz and Founders Fund with $752 million raised and a $2.9 billion valuation as of 2022, built its model in oil and gas, placing 30,000 contractors by 2021. It acquired traditional staffing agencies, aggregated their contractor pools into a single marketplace, and layered workforce management software on top. Workers get a profile. Companies get a dashboard. The platform takes a cut before depositing the check.
Buildforce, a seed-stage startup out of Austin, focuses specifically on construction workforce matching with AI-driven recommendations. They acquired Ladder, an Atlanta staffing tech company, in December 2024 and are pitching themselves as the layer that centralizes workforce data and visualizes labor trends. Bridgit Bench occupies the planning tier, helping GCs allocate people across multiple projects with forecasting tools. Procore added workforce management to its construction software suite, bringing AI-enabled labor scheduling to its existing user base.
Add GigSmart (temporary construction labor through a gig app) to the mix and you have an ecosystem.
What the Worker Loses
Traditional construction staffing agencies mark up labor 35 to 60 percent above the worker's base pay, according to TempGuru's 2026 staffing cost analysis. That sounds like robbery until you understand what the markup covers: employer-side FICA (7.65%), workers' compensation insurance (2 to 8% depending on role and state), unemployment insurance, and general liability. The compliance layer alone accounts for 12 to 22 percent of the rate before any agency margin.
Gig platforms quote lower. The rates look better on the app screen. But as TempGuru's analysis bluntly states: "The gig model doesn't eliminate employer obligations. It strips them from the platform's ledger and places them on yours." When a platform classifies a framer as a 1099 independent contractor instead of a W-2 employee, there is no workers' comp. No unemployment insurance. No employer FICA contribution. No paid sick leave.
Run the numbers. A worker earning $36.54/hour through a W-2 arrangement actually receives a total compensation package worth roughly $41 to $45 per hour when employer-funded benefits are included. A 1099 gig worker earning $39/hour on the app is, after self-employment taxes (15.3% FICA) and the cost of individual health and liability coverage, taking home less than the W-2 worker earning $36. The premium on the screen is an illusion.
| Cost Component | W-2 Employee | 1099 Gig Worker |
|---|---|---|
| Nominal hourly rate | $36.54 | $39.00 |
| Employer FICA (7.65%) | Paid by employer | Worker pays full 15.3% |
| Workers' compensation | Employer-funded | None (or self-purchased) |
| Unemployment insurance | Employer-funded | Not eligible |
| Health benefits | Often employer-subsidized | Self-purchased |
| Effective hourly value | ~$42-45 (total comp) | ~$31-33 (after taxes + self-coverage) |
Effective hourly value for 1099 workers calculated using self-employment tax (15.3%), estimated individual health insurance ($450/month per KFF 2024 employer survey benchmark), and basic general liability coverage ($150/month). Actual figures vary by state, trade, and individual circumstances.
Rating Systems Nobody Audits
Instawork's AI "improves with every shift to verify skills and fit." Workrise builds contractor profiles with ratings for previous work. Every platform tracks punctuality, completion rates, client satisfaction scores.
None of these systems have been independently audited for bias.
Human Rights Watch published "The Gig Trap" in May 2025, documenting how platform rating systems in the broader gig economy lead to opaque deactivations, discriminatory outcomes, and earnings manipulation. Workers interviewed described being cut off from the platform with no explanation, no appeal process, and no way to transfer their work history to a competitor. The UC Berkeley Labor Center has recommended banning fully automated deactivations and ratings-based work allocation systems that lack transparency.
Construction work adds a dimension these reports barely touch. If an Uber driver gets deactivated unfairly, they lose income. If a plumber gets deactivated from a construction staffing platform mid-project, a homeowner's bathroom renovation stalls. The downstream effects ripple into project timelines, contractor penalties, and construction loan draw schedules. Nobody in the platform ecosystem has grappled with what happens when algorithmic workforce management meets the physical reality of a half-finished house.
In Defense of the App
I need to be honest about the counterargument because it is strong.
A skilled electrician in Houston who can pull $45/hour shifts on demand through Instawork may genuinely prefer that arrangement to waiting for a union hall call, hoping a single general contractor has consistent work, or being at the mercy of a staffing agency that takes 50 cents of every dollar billed. Platform portability means your reputation travels with you. Twenty-four-hour shift filling means less downtime between jobs. Real-time credential verification means fewer unqualified workers showing up on job sites, which is a safety gain.
And the matching gap is real. If these platforms get even 10% of those 96,400 monthly mismatched workers into the right jobs, the value is enormous, not just in aggregate economics but in real paychecks for real people who were sitting idle in the wrong city. The traditional system, with its fragmented agencies and phone-call networks, was not solving this problem. Something needed to change.
The question is not whether platforms improve matching speed. They clearly do. The question is whether speed can coexist with the worker protections that took organized labor a century to build.
What the DOL Is Doing About It
The Department of Labor proposed in 2025 to replace its 2024 independent contractor rule with a clearer test that narrows the definition of who qualifies as a genuine 1099 worker. The DOL's own projections suggest the new rule could enable 250,000 to 750,000 new independent contractor arrangements, not by reclassifying employees, but by reducing the legal ambiguity that makes companies afraid to use contractors at all.
Annual cost savings from the proposed clarity: $682.7 million. One-time familiarization costs: $488.2 million. The DOL explicitly criticized the 2024 rule's "integral part" test as so vague that virtually all work could be deemed integral, pushing employers to classify genuine independents as employees just to avoid lawsuits.
Meanwhile, the EEOC settled its first AI-discrimination lawsuit against iTutorGroup for $365,000 after the company's hiring algorithm automatically rejected older applicants. The settlement was small. The signal was not. The EEOC has stated its intent to enforce anti-discrimination law wherever AI makes employment decisions. Construction labor platforms that use algorithmic scoring to rank, select, and deactivate workers are squarely within that enforcement perimeter.
What This Means If You Are Building a Home
If your general contractor is staffing your project through a platform, the workers showing up tomorrow morning were selected by an algorithm you cannot see, rated by a system nobody audits, and classified as independent contractors who carry no workers' compensation from the hiring platform. If one of them falls off your roof, the liability chain gets complicated fast.
Ask your GC: are the workers on my site W-2 employees or 1099 contractors? If 1099, who carries workers' comp? Who carries general liability? If the answer involves a platform app, verify that the platform or the worker actually holds current coverage. In states like California, where misclassification penalties are steep, a contractor staffing through gig platforms without proper coverage is exposing both themselves and you.
None of this means platform-dispatched workers do bad work. Many are highly skilled tradespeople who chose flexibility over a single employer. But the system around them, the insurance, the accountability, the legal classification, is still catching up to the technology.
Limitations
Workrise revenue and valuation figures are estimates from Sacra, a private-company research firm, not company-disclosed numbers. Instawork's 8 million workers span all industries; no public data breaks out construction-specific placements. The HRW "Gig Trap" report focused on ride-hailing and delivery platforms; construction gig dynamics differ in meaningful ways (higher skill floor, different injury profile, project-based rather than shift-based work). My $5.4 billion matching gap calculation assumes a static 20% closure rate and linear fill dynamics, which oversimplifies a labor market with regional, seasonal, and trade-specific variation. No longitudinal data exists comparing career outcomes (wages, injury rates, retirement savings) for platform-dispatched construction workers versus traditionally employed ones. That study needs to happen.
Marcus Washington covers workforce and labor for AI Home Building. His father spent thirty-one years in the Ironworkers. He had a pension, a health plan, and a foreman who knew his name. The app does not offer those things.