Half the people you hire don't work out. You've lived this for 20 years, and you keep running the same process. Resume + gut feel is a habit, not a strategy. Here's why it fails and what replaces it.
The math of a bad hire
Industry numbers put trades-tech turnover inside the first year at 45-55%. Replacement cost runs 50-200% of annual compensation when you factor in lost productivity, training, callbacks, and customer attrition. On a $65k/year tech, a single mis-hire that washes out at month six is $40k+ of margin you'll never get back. You eat that quietly because it's spread across the P&L, but it compounds.
Why the resume lies
A resume tells you what the candidate did. It tells you nothing about how they did it, whether their last shop bailed them out, whether they'll synthesize on the truck or freeze. The information you actually need to predict performance - clarity, prioritization, taste, customer empathy - isn't on the resume because no candidate writes a resume against those axes. You'd have to manufacture the signal yourself.
Worse: the better the candidate is at writing resumes, the weaker the correlation with how they'll perform. The cover- letter tech-bro looks great on paper and chokes on the synthesis prompt.
Why gut feel lies
Gut feel after a 30-minute interview is a measurement of three things: whether you liked the candidate, whether they presented well, and whether they handled small talk smoothly. None of these predict whether they'll be a great tech on a hot July afternoon with a frustrated customer and an ambiguous diagnostic. Your gut got trained on signals that worked in 2005 when the labor pool was different and the AI multiplier didn't exist.
What replaces it
Predictive hiring replaces it. You decide what outcome you're hiring for - revenue per tech, callback rate, tenure, time to proficiency - and then you build a process designed to predict those outcomes. The 5-Filter system is one such process; it is not the only one, but it is the only one I've found that works on trades roles in particular. Read the walkthrough, then read the long-form.
Whatever you do, stop optimizing your hiring process for "do I like this person?" You've been doing that for two decades and half your hires have not worked out. Try the other thing.