What I learned about hiring from 20 years of getting it wrong and right

Every bad hire I ever made had warning signs I explained away. Every great hire I almost passed on for reasons that look stupid in hindsight. After 20 years building teams in commercial roofing and sheet metal, the pattern is so obvious it embarrasses me.

TL;DR: After 20 years building teams in construction and roofing, my hiring framework is simpler than it used to be. I care about three things. Do they do what they say. Do they get better without being told. Are they honest when something goes sideways. Everything else can be taught. Those three cannot.

The three signals that actually matter

I used to interview for the wrong things. Years in the trade. Brand-name employers. A clean resume with no gaps. None of that predicted who would be on a roof at 6 a.m. solving a problem the right way.

Now I screen for three things, in this order.

One. Do they do what they say. Sounds basic. It is not. I watch the whole interview process for it. If they say they will send references by Tuesday, I am tracking Tuesday. If they tell me they are available for a second conversation at 9 a.m. Thursday, I am watching for whether they confirm. The interview process is a sample of how they will work for me. If they ghost on a small commitment now, they will ghost on a bigger one when they are on payroll.

Two. Do they get better without being told. I ask about the last thing they learned on their own. Not a class their employer paid for. Something they decided was a gap and closed. The answer tells me whether I am hiring someone who needs a babysitter or someone who runs their own development. In a business where the products change every two years and the codes change every three, the second person is worth three of the first.

Three. Are they honest when something goes sideways. I ask about the worst mistake they made on a job and what they did about it. If the answer is some version of “it was not really my fault,” the interview is over. I have spent a career cleaning up messes made by people who would not admit the mess existed. I am not doing that again.

Everything else can be taught. EPDM, TPO, ANSI standards, takeoffs, CRM hygiene, how to talk to a property manager. All teachable. The three above are not.

Every bad hire I made had warning signs I explained away

I want to be specific about this because most hiring post-mortems lie.

The bad hires were not surprises. The warning signs showed up in the first two conversations. I just had a need, and the need rewrote my judgment.

Examples of the rationalizations I have personally used:

• He was late to the second interview, but traffic is real.
• She bad-mouthed her last three bosses, but those companies do have reputations.
• His references were all friends, not managers, but he said his old manager retired.
• She could not explain why she left her last job in 90 seconds or less, but she is nervous.
• He kept saying “to be honest with you,” but it is just a verbal tic.

Every one of those rationalizations cost me money. The traffic one cost me about $180,000 in lost productivity and a project I had to personally rescue. The verbal tic one cost me a client.

The lesson is not that any one of those signs is fatal on its own. The lesson is that when I notice three of them in a single candidate and still hire, I am the problem, not them. I needed the seat filled and I let urgency override pattern recognition. That is the most expensive mistake an executive can make in a tight labor market.

Every great hire I almost passed on

The other side of this is harder to talk about because it requires admitting bias.

The best estimator I ever hired came in wearing the wrong clothes. The best project manager I ever hired had a resume gap I could not initially explain, because she had been caring for a sick parent and did not want to lead with it. The best superintendent I ever promoted spoke English as a second language and I almost dropped him from the pool because his cover letter had grammar errors. He could read a set of drawings faster than anyone I have worked with in 20 years.

The pattern in my misses. I was screening for polish when I should have been screening for capability. Polish is a proxy for confidence in middle-class professional norms. Capability is the actual asset. They are not the same thing and they often live in different people.

I have a rule now. Before I reject a candidate, I write down the specific reason in one sentence. If that sentence reads back to me as a reason about how they present rather than how they work, I keep them in the pool and dig further.

How to interview for character without sounding like an HR poster

You cannot ask people if they have character. They will say yes. So will the people who do not.

What works is asking for stories that put them on the wrong side of an easy lie. Here are the four questions I use, and what I am actually listening for.

Tell me about a time you broke something at work and had to tell someone. I want a specific incident, a specific person they told, and a specific outcome. Vague answers mean either they have never owned a real mistake or they are hiding one.

What did your last manager get wrong about you? The answer tells me whether they can name a real gap between how they see themselves and how they get read. People who cannot answer this either do not reflect or are performing for the room. Both are problems.

What is the last thing you taught yourself outside of work? Not their employer’s required training. Something they decided to learn. A YouTube series, a side project, a certification they paid for personally. If nothing comes to mind, they are not curious. In a trade that changes this fast, curious beats credentialed.

If I called the person you worked next to every day, not your boss, what would they say is hard about working with you? Anyone who says “nothing” or “I push too hard” is reading from a script. Anyone who answers honestly is showing you the self-awareness you are buying.

None of these are clever or original. What matters is that I follow up. I ask for the next layer of detail until the answer is either specific or visibly invented. That is where character shows up.

The mistakes I see other construction executives keep making

I have built four teams across four companies. I have also talked to a lot of peers. The mistakes are remarkably consistent.

Hiring for the resume instead of the job. A guy with 15 years at a competitor is not automatically better than a sharp five-year operator who has worked across three different environments. The 15-year resume tells you he survived. It does not tell you he grew. Ask what changed about how he works between year one and year fifteen. If the answer is nothing, that is your answer.

Letting the urgency of an open seat set the bar. Every time I have lowered my standard because I needed a hire by the end of the month, I have paid for it for the next 18 months. Open seats are expensive. Bad hires are far more expensive. Run the math the next time you are tempted.

Outsourcing the gut check. If a candidate has gotten through five interviews and your stomach is still uneasy, do not hire them because everyone else liked them. You are the one who has to live with this person on your team. Trust the signal. I have ignored that signal three times in my career. Three for three on regret.

Confusing cultural fit with cultural sameness. Cultural fit is whether someone shares your values around quality, accountability, and how you treat people. It is not whether they went to the same kind of school or watch the same shows. The first is a real screen. The second is how you build a team that all thinks the same way and misses the same things.

Treating reference checks like a formality. A real reference call lasts 25 minutes and asks specific questions. Would you hire them again. What would the person who sat next to them say. What is the one thing they need to work on. If the reference dodges the third question, you have an answer.

Cultural fit without discrimination

This deserves its own section because it is where good intentions get people in trouble.

Cultural fit, done wrong, becomes a coded screen for race, age, class, accent, gender, or background. That is illegal and it is also bad business. The people you exclude are the people whose perspective would have caught the mistake your team was about to make.

The fix is to be specific about what your culture actually requires. Not vibes. Behaviors.

My culture requires that people tell the truth when a job is in trouble. It requires that people follow up on their commitments without being chased. It requires that people treat the guys on the roof with the same respect as the people in the boardroom. Those are testable behaviors. I can screen for them in an interview and watch for them in a 90-day evaluation.

If your description of “culture fit” cannot be turned into specific behaviors, it is not a culture screen. It is a comfort screen. Comfort screens are how mediocre teams stay mediocre.

What to actually do on Monday

If you are sitting on an open seat right now, here is the practical version.

1. Write down the three behaviors this role actually requires. Not credentials. Behaviors. “Closes the loop on commitments within 24 hours.” “Identifies their own training gaps.” “Owns mistakes in writing.”
2. Build your interview around evidence of those behaviors. Past stories, not hypotheticals. Specifics, not summaries.
3. Make the interview process itself a behavior test. Track who confirms, who follows up, who sends thank-you notes that show they were actually listening.
4. Talk to a real reference. Not a friend. The person who managed them or sat next to them. Ask the three questions above.
5. If your stomach is uneasy after all of that, do not hire. The cost of waiting is always less than the cost of a bad hire.

The framework is not complicated. It is just hard to follow when you are tired and short-handed. Which is exactly when it matters most.

One more thing

The best hire I ever made told me in the interview that she had been fired from her job before the one she was leaving. She walked me through what happened, what she learned, and what she had done differently since. No drama, no spin, no scapegoats. I knew within five minutes she was the person.

She ran a $14 million book of business for me for six years and never lost a key account. The firing was a non-event. The honesty about it was the whole signal.

If you’re leading a team and want to compare notes on building one, get in touch.