The hiring problem we keep solving wrong

Authored by: Gregory Hair

Last month I read a stat that stopped me cold. For the first time in fifty years, US college grads have a higher unemployment rate than trade school grads. Only 30 percent of the 2025 college class found full-time work in their field. Meanwhile, new construction hires pull in $48,000 starting wages while new professional services hires sit at $39,500.

Something has shifted, and most of us in frontline industries are still pretending it hasn’t.

For years we told ourselves a story about labor shortages. Workers don’t want to work. Gen Z is lazy. The generation before them ruined everything. None of that holds up against the numbers. There are 7.4 million open jobs in the US right now, but only 5.2 million hires happening. So that 2.2 million gap is not a worker problem. It is a process problem, and we built it ourselves.

The application black hole

Here is something nobody wants to admit. Roughly one in five postings on the biggest hiring platforms is a ghost job. On LinkedIn the rate climbs to 27 percent, with construction leading every other industry. Companies post jobs they have no near-term intent to fill, then act surprised when applicants stop taking them seriously.

Then there is the funnel itself. iCIMS surveyed frontline hiring managers last year and 91 percent said their roles were urgent. Yet 60 percent of frontline workers said they had abandoned an application because the form was too long. Sixty-one percent of US job seekers got ghosted after an interview in 2024, up nine points in six months.

So the hiring manager thinks the workers are flaky. The workers think the company is flaky. Both are kind of right, and neither is getting hired.

Now the places solving this are not doing anything fancy. Chipotle cut time-to-hire from 12 days to 4 by letting candidates apply by text and skipping resumes entirely. Compass Group hires 120,000 people a year with 20 recruiters. The trick was not AI magic. It was removing friction nobody needed in the first place.

The accidental manager tax

But here is what I keep coming back to. Even if you fix the front door, the back door is wide open. Gallup’s research shows managers account for at least 70 percent of the variance in team engagement. And then this: 82 percent of new managers in the UK get zero formal training. Half of all working managers hold no qualification at all.

We promote our best welder, our best driver, our best installer, and tell them to manage people now. Then we wonder why they grind out the team and burn out themselves. McKinsey found frontline managers in retail and logistics typically oversee 20 to 30 reports. Their white-collar peers manage 6 to 10. In some retail districts, store managers spend as little as 10 minutes a day coaching anyone.

So this is not a soft problem. The OECD estimates the UK has 2.4 million accidental managers costing employers around 84 billion pounds a year. Half the productivity gap between the UK and US traces back to management capability. We bleed money out of our businesses because we will not train the people we already promoted.

What the data shows works

Now here is the part that gets me. Companies spend on average $9,100 per employee per year on software. They spend $1,200 per employee per year on training. Read that again. Eight times more on tools than on the humans using them.

Yet the operators getting this right look almost boring from the outside. Costco pays a median wage around $25 an hour against a retail average of $17. Sixty percent of their staff have been there five years or longer. They promote almost every store manager from within. Sam’s Club tried a similar approach in 2018 and saw productivity climb. Trader Joe’s, In-N-Out, QuikTrip, all running the same playbook. Pay above market. Train first-line supervisors. Stabilize schedules. Cross-train. Carry slack in the staffing model.

Sounds obvious. It is not. When Walmart raised starting wages in 2015, the stock dropped 10 percent the same day. The market literally punished them for paying frontline workers more. So that trading day explains why most operators default to squeezing labor costs. Every shred of evidence says it backfires anyway.

Where this leaves us

So I keep telling people in our industry the same thing. Stop blaming workers for not wanting the job. Stop expecting AI to solve a problem that is mostly about how we treat the humans we already have. The demographics are locked in. By 2031, 41 percent of the construction workforce will retire. Japan will be short 11 million workers by 2040 even after a generation of pro-natalist policy. You cannot recruit your way out of this. You cannot automate your way out either. McDonald’s killed its AI drive-thru pilot. Walmart pulled its shelf-scanning robots. Tesla scrapped its Model 3 conveyor automation, with Musk admitting humans were underrated.

So the boring answer keeps winning. Pay people well. Train the people who manage them. Make the application take five minutes instead of fifty. Treat the frontline like the business depends on them, because it does.

We called them essential. Then we underpaid them, undertrained them, undermanaged them, and now we cannot replace them. That is the entire frontline labor story right now. Until we own that, nothing we do at the hiring level is going to stick.

Author Bio: Gregory Hair, Owner, Landscaper, SLIDE Living