Retention in roofing: why the third year is when you actually lose people

First-year turnover is the noise. Third-year departures are the signal. If you run a roofing company and you keep getting blindsided when your best foreman, estimator, or project manager walks out the door at month 32, the problem isn’t loyalty. It’s your raise schedule, your career conversations, and the math you haven’t done.

TL;DR: First-year turnover is loud. Third-year departures are the ones that cost you. By year three an employee knows what they’re worth on the open market, and your raise schedule almost never matches. Read the curve, have a real career conversation in months 24 to 30, and act before they interview elsewhere.

Year one turnover is a hiring problem, not a retention problem

When someone quits in the first six months, that is a hiring miss. Wrong fit, wrong expectations, wrong manager, wrong commute. You can fix it by tightening your interview process and your first ninety days. That work matters, but it is not what this post is about.

The reason year one turnover gets all the attention is that it’s loud. New hires quit dramatically. They no-call no-show. They leave a voicemail. They post about it on Facebook. HR sees the ticket, the GM sees the empty truck, and everyone files it under retention. It is not. It is recruiting and onboarding.

Real retention is what happens to the people who already proved they can do the job. The ones you trained, certified, and trusted on the biggest accounts. The ones whose names show up on every bid you win. Those people don’t leave loudly. They leave quietly, with two weeks notice and a Friday goodbye lunch. By then it’s too late.

Why the third year is the breaking point

By month 30 a roofing employee has three things they didn’t have on day one. They know your customers. They know your margins. And they know what they’re worth on the open market. That last one is the killer.

Here is the curve I’ve watched for twenty years. Year one, they are learning. They are grateful for the job and the training. Year two, they are producing. They are billable, accountable, and starting to make your company money. Year three, they are valuable. They run jobs without supervision. They train the new hires. They are the person a customer calls when something goes sideways at 7pm on a Thursday.

That third-year person costs you sixty to eighty thousand a year in salary. On the open market they’re worth ninety to a hundred and ten. Your raise schedule, the one HR built five years ago, gives them three percent. The competitor across town offers them fifteen. Math wins.

The brutal part is that nothing has gone wrong. They like you. They like the team. They like the work. They just got an offer they can’t ignore, and your last career conversation with them was during their hiring interview.

How to read the curve before they leave

People who are about to quit leave a trail. You just have to know what to look for. None of these signals are proof on their own. Two or three together, and you have about ninety days to act.

The first signal is calendar. They start taking PTO they never used before. Friday off. Tuesday afternoon. Vague reasons. That is interview prep, second-round interviews, or pre-employment screening. Anyone in trades knows interviews happen during the workday because everyone else is on a jobsite too.

The second signal is LinkedIn. They suddenly update their profile, add certifications they have been quiet about for a year, or start connecting with recruiters. You don’t need to stalk them. You just need to notice when someone who never posts starts posting.

The third signal is engagement. They stop volunteering for the optional things. The Saturday walk of a difficult roof. The lunch with the new hire. The phone call to the customer who needs hand-holding. They are still doing their job, but they have stopped doing the extra.

The fourth signal is questions. They start asking about your 401k vesting schedule, when their next stock or bonus payment hits, or how PTO payout works at termination. That is not curiosity. That is due diligence.

The fifth signal is silence in the one-on-one. The person who used to argue with you about scheduling, pricing, or estimating goes quiet. They nod. They agree. They no longer push back. Disengagement looks a lot like cooperation if you are not paying attention.

The conversation that prevents 80 percent of third-year departures

Somewhere between month 24 and month 30, every employee you want to keep needs a real career conversation. Not a performance review. Not an annual. A career conversation. There is a difference, and most companies in this industry have never run one.

I block ninety minutes. I leave the building. I buy lunch or coffee somewhere neutral. And I ask four questions. The fourth one is the one that matters.

One. Where do you want to be in three years? Title, money, scope. Not corporate-speak. Real numbers and real responsibilities.

Two. What do you need to learn to get there? Certifications, licenses, exposure to parts of the business you haven’t touched yet, mentorship.

Three. What is the biggest thing slowing you down right now? In this company, in this role, with me as your boss. Honest answers only.

Four. If you were going to leave us in the next year, what would the reason be? This is the question that breaks the room open. Nobody asks it. Everybody answers it.

The answer to question four tells you exactly what to fix. Sometimes it is money. Often it is not. It is a manager problem, a growth ceiling, a commute, a benefits gap, a feeling that they have been forgotten by leadership. You can’t fix what you don’t know, and they will not volunteer it in an annual review with a form to fill out.

Run this conversation before you think you need to. Once they are already interviewing, you are negotiating against a deadline you can’t see.

The real cost of replacing a third-year employee

Most roofing companies undercount this number by half. The Society for Human Resource Management benchmarks replacement cost at six to nine months of salary for skilled roles. In trades, with safety training and licensing on top, it runs higher. Here is the breakdown for a project manager making $85,000.

Recruiting fees or internal recruiter time. Five to twelve thousand if you use an agency. Two to four thousand in internal time if you do not.

Vacancy cost. The work does not stop. Either a peer covers it on overtime or a job slips. Conservative estimate, fifteen to twenty thousand over the ninety days you are open.

Onboarding and ramp. The new hire is at fifty percent productivity for the first ninety days and seventy-five percent for the next ninety. That is roughly seventeen thousand in lost output on an $85k base.

Training and certifications. OSHA, manufacturer certs, software, safety. Two to five thousand depending on what your bench needs.

Lost institutional knowledge. The relationships with three repeat customers, the workaround for the difficult inspector in Brookfield, the supplier rep who answers their cell on Sunday. You can’t price this until you lose it, but on the next bad job it shows up as a five-figure mistake.

All in, replacing an $85k project manager costs you somewhere between $60,000 and $110,000. A $25,000 raise to keep someone in year three is not generosity. It is math.

When a counter-offer is the right move, and when it is not

The conventional wisdom says never counter-offer. The conventional wisdom is wrong. It is wrong because it treats every departure the same, when the right answer depends entirely on why they are leaving.

Counter when the issue is purely market comp and they brought you the offer before they accepted. That is a person who wants to stay and is giving you a chance to keep them. Match it, beat it, and write a letter committing to a structured comp review every twelve months so they don’t have to force your hand next year.

Do not counter when they have already accepted or when the issue is not money. If they are leaving because the manager is the problem, the commute is the problem, or the work itself is the problem, money buys you six months and then they leave anyway with a worse story about your company. Pay them out professionally and let them go.

The harder rule is this. If you only find the money when they threaten to quit, you’ve taught your entire team that the path to a raise is interviewing somewhere else. That is a culture you don’t want. The fix is to pay the market before they have to ask. That requires comp benchmarking, not loyalty speeches.

Practical takeaways

Five things you can do this quarter to fix your third-year retention problem.

One. Pull a tenure report. Sort your employees by start date. Identify everyone hitting month 24 in the next six months. Those are your conversations.

Two. Benchmark comp. Use BLS data, NRCA salary surveys, or a sample of three competitor job postings. Find out where your roles actually sit against the market. Do not guess.

Three. Calendar career conversations. Ninety minutes, off site, the four questions above. Run them before you think you need to.

Four. Build a real raise schedule. Not the three percent annual you copy-pasted from your handbook. Tie raises to scope expansion, certifications earned, and market movement. Tell people the formula so they stop guessing.

Five. Track quit signals weekly. PTO patterns, LinkedIn activity, engagement drop, comp questions, one-on-one silence. Put it on the same scoreboard as your safety and revenue numbers.

None of this is complicated. It is just work most operators are too busy to do until the day someone walks in with a resignation letter, and then they pay ten times the cost to fix it after the fact.

The honest truth about retention

Companies don’t lose great people because the great people are disloyal. They lose them because nobody in leadership did the math, had the conversation, or paid attention to the signs. The information was always there. The will to act on it was not.

If you run a roofing or construction business and you are tired of watching your best people walk out the door to your competitors, the fix is not a ping pong table or a culture deck. It is a calendar reminder at month 24, a number on a spreadsheet, and ninety honest minutes off site.

I’ve run those conversations for two decades. I’ve lost people I should have kept, and I’ve kept people I almost lost. The difference was almost never charisma. It was knowing when to ask and being willing to hear the answer.

If you want me to walk your leadership team through what a real retention system looks like in roofing and construction, you can book me to speak at your next leadership meeting or industry event. I’ll bring the spreadsheets.