Fraser Patterson, Founder & CEO, January 2026

Once we fix access to skilled construction workers, fixing the shortage itself becomes our primary focus.

Construction is indeed losing workers faster than it’s replacing them with retirements, burnout, injury and career switches outpacing successful new entrants by roughly 80,000 workers per year. This puts the industry on track, to lose ~800,000 workers over the next decade under the status quo, even before accounting for growth in construction demand.

This shortfall is often explained away as a lack of interest in physical work from younger generations who rather then learn a durable skilled craft would apparently rather spend their days staring at a blue rectangle all day. But that explanation explodes on inspection.

Interest exists at the entry point

The problem isn’t actually interest - each year, roughly 300,000–500,000 people attempt to enter construction through apprenticeships, trade schools, CTE programs, helpers, and entry-level roles. Union apprenticeship programs alone routinely receive 2–5x more applicants than available seats, and high-school trade enrollment has remained stable or grown in many regions.

The problem is that fewer than ~40% of those entrants persist long enough to become economically productive skilled workers. More than half leak out of the system before productivity, not because they lack interest, but because the early path is very fragile with structural and economic failure points.

Most skilled trades require 3–5 years to reach full productivity and during that time, employers cannot justify paying full wages before output arrives, while workers face higher physical demands, unstable hours, upfront costs (tools, transportation, certifications), and income volatility. On paper, a first-year apprentice earns about $43K on average, compared to roughly $29K in retail but in practice, upfront costs and instability more often that not erase that premium. Add to that a single bad placement, stalled project, or short-term financial shock and frequently a permanent exit ensues.

Crucially, today’s system treats early failure as terminal because there’s no mechanism to reassign workers who land in the wrong role or at the wrong time, and no way to buffer early economic risk. Each employer bears risk in isolation, and each worker absorbs failure personally.

Act 2 changes that system

As Skillit scales, our AI hiring infrastructure is becoming the system of record for real-time labor demand — capturing where work is happening, which roles are needed, when they’re needed, what workers are paid for which skills, and which pathways lead to the fastest route to economic productivity. This isn’t static market data; it’s live signal generated inside actual hiring workflows.

That demand data allows Skillit to expand the role of its AI agents beyond hiring coordination into workforce formation. And because Skillit sits directly inside sourcing, interviewing, offers, compliance, onboarding, and placement, early failure can be treated as a routing problem, not an exit. When a placement doesn’t work, workers are rapidly reassigned into other verified, active demand instead of being lost from the industry entirely. Idle time is minimized through sequenced placements, and mismatches are corrected early before they compound into attrition.

Over time, this same infrastructure lets us move further upstream. We can point our labor acquisition engine toward high schools, early-career entrants, second chance citizens etc. (aka people with the potential to succeed in the trades) and not just the highly skilled, job-ready workers we focus on today.

At the same time, Skillit can enable pooled economic risk across a large network of employers and workers. Rather than asking any single employer to fully underwrite the low-productivity years of an apprentice, early-career risk is spread across many hires, companies, regions, and timelines. We could also compress the cost of becoming a productive worker by using our purchasing power to lower tool costs or eliminate them entirely.

This makes it possible to stabilize early earnings, offset upfront costs, and smooth income volatility during the learning ramp, all without having to pay wages disconnected from productivity. The goal being not to shorten the path to craft mastery (a possible positive side effect for sure), but rather to help more people survive the path to it.