In a crew of ten, one person leaving is a tenth of the company walking out the door. Most of what makes the other nine stay gets decided in the first ninety days on the job and in how the boss runs the week after that.

What turnover actually costs a ten-person crew

I want to start with the number, because owners run on numbers before they run on feelings. The Center for American Progress reviewed 22 case studies covering jobs paying under $75,000 a year in the United States and found "a typical cost of turnover of 20 percent of salary." Gallup, also writing about the US, puts the range wider: the cost of replacing an employee "can range from one-half to two times the employee's annual salary." Both are US figures. Neither is specific to a ten-person crew. Here is what they mean once I put a real Manitoba wage against them.

Job Bank's own wage report for a general construction labourer in Manitoba puts the median at $22.00 an hour, updated November 19, 2025. At a 40-hour week, that is $45,760 a year. Apply the CAP low end and one departure costs roughly $9,150. Apply Gallup's wider range and it runs from about $22,900 up to $91,500. Most of that is never on an invoice. It is the overtime the other nine people pick up, the weeks the shop runs short-handed, the estimate that used to live in one person's head and now does not, and the weeks a new hire spends being slower than the person they replaced. In a crew of ten, that is not a rounding error. It is a real percentage of the whole company's output, gone, for months.

Why people leave small crews

Gallup has studied this at a scale I never will. Its research on the American manager found that "one in two employees have left a job to get away from a manager and improve their overall life at some point in their career." That is not a rare, made-up-boss problem. It is close to the single most common reason people quit.

In my own experience running crews before I did this for a living, it showed up the same four ways almost every time. Unclear expectations: someone was handed a job, not a job description, and only found out where the line was after they had already crossed it. No feedback: a good six months could pass with nothing from me except an annual review, or worse, nothing until a blow-up. Pay that quietly drifted below what the same role paid elsewhere in town, because nobody checked. And me, the owner, as the bottleneck: decisions sitting on my desk while the crew waited to hear back on something I could have answered in five minutes.

Hiring for fit in the first place

None of what follows matters much if the person was wrong for the seat on day one. That starts at the interview, not at the exit interview. I wrote about how I actually run a first interview in interviewing a candidate for the first time, and about matching how someone is wired to the actual role in which DISC profile fits a sales role. Get the fit right at the door and everything below gets a lot easier. Get it wrong, and you can read what that costs in what a bad hire costs a Manitoba business.

The first ninety days

In my experience, whether someone is still there at the one-year mark gets decided in this window, not later. I use four checkpoints, and I treat each one as a thing that has to happen, not something that happens on its own.

WhenWhat happensWhy it matters
Day 1A workstation, tools, and paperwork ready before they arrive; a name for who they report to and who they sit with at lunchNobody should spend their first morning waiting for someone to figure out where they go
Week 1One task they can finish and see the result of, plus a short check-in on how the week wentA first win tells them they can actually do this job, before doubt has time to set in
Month 1A real conversation on how the role compares to what was described, and what is and is not working yetThis is the last easy chance to fix a mismatch before it hardens into a habit on either side
Month 3A formal review against the job itself, not a gut feeling, and a clear yes on keeping them past probationWaiting past this point to decide costs more than it saves; see the probation post below for what Manitoba law actually requires here

The four checkpoints do not change by role. Only what is being checked changes.

The weekly habits that keep people

Past the first ninety days, in my experience it is a small set of weekly habits that keeps a crew together, not a program. A fifteen-minute one-on-one, every week, with whoever they report to, so small problems get named while they are still small. A scoreboard everyone can see, so people know how the crew is doing without having to ask. And saying thank you in public, correcting in private, every time, so people are not guessing which one is coming when you pull them aside.

Pay and the market

I check Job Bank's median wage for each role I run once a year, not once and never again. Wages move, and a crew that has not had a raise while the market moved past them will notice before you do. Statistics Canada's Job Vacancy and Wage Survey gives a sense of how tight the market is right now: in the first quarter of 2026, Manitoba employers had 17,810 job vacancies open, a vacancy rate of 2.8 percent, at an average offered wage of $26.55 an hour across all industries. That is not a wage for any one trade, it is the whole province, but it is a useful check on whether your own numbers have kept up.

When to let someone go anyway

Sometimes the honest answer is that the fit was wrong from the start, and no amount of one-on-ones fixes that. I would rather make that call inside the first ninety days than carry it for a year and let the rest of the crew watch me avoid it. What Manitoba law actually requires around probation and notice is its own subject, and I do not repeat it here; I cover it properly in what Manitoba law actually says about a probation period. Read that before you act, not after.

The three ways I can help, cheapest first

Start at the cheapest one that solves your problem.

Do it yourself on Quadrant. Quadrant, the DISC assessment and team tool I built, runs a DISC assessment for your whole crew and every applicant, and makes sure everyone gets their one-on-one every week with the person they report to. It also runs scheduled review cycles where both sides rate the same job-description lines, for the cost of a subscription instead of a fee.

Hiring Services, by the job or by the month. If the turnover problem starts at hiring, I write the posting, screen every applicant on Quadrant, and hand you a ranked shortlist with an interview guide. One Hire is $2,999 per role. Always Hiring is $1,500 a month for up to three roles, or $2,999 a month for up to ten, three-month minimum, then month to month. Full details on the Hiring Services page.

Contract Recruiting, the full search. You give me the finished job description, I run the whole search end to end, and you stay as involved as you want. That is 15% of first-year salary, $5,000 minimum per role, charged once, no retainer and no subscription. Full details on the Contract Recruiting page.

Questions owners ask first

How much does losing one person actually cost a small crew?

The Center for American Progress reviewed 22 case studies on jobs paying under $75,000 a year in the US and found a typical turnover cost of 20 percent of salary. Gallup puts the range wider: replacing one employee can run from one-half to two times their annual salary, also a US figure. Applied to Job Bank's Manitoba median wage for a general construction labourer, $22.00 an hour, that is somewhere between about $9,150 and $91,500 for one departure, most of it hidden in overtime, slower jobs, and lost knowledge rather than on an invoice.

Is a bad manager really the top reason people quit?

It is close to it. Gallup's research on the American manager found that one in two employees have left a job to get away from a manager and improve their overall life at some point in their career. In my own experience running crews, the specific pattern is almost always the same: unclear expectations, no regular feedback, pay that quietly drifted below market, or too many decisions stuck waiting on the owner.

What should happen in a new hire's first ninety days?

In my experience it comes down to four checkpoints: day one, week one, month one, and month three, each with a specific thing that has to happen rather than being left to chance. I lay out the checklist I use in the first ninety days section above. Most of what determines whether someone stays past a year gets decided in this window, not later.

How often should I check whether my pay is still competitive?

Once a year at minimum, against Job Bank's own wage report for the occupation and province, which Job Bank updates roughly annually. Statistics Canada's Job Vacancy and Wage Survey is worth a glance too: in the first quarter of 2026, Manitoba employers had 17,810 job vacancies open, a vacancy rate of 2.8 percent, at an average offered wage of $26.55 an hour across all industries. If your crew's pay has not moved and that average has, you are quietly falling behind.

When should I let someone go instead of trying to fix it?

When the fit was wrong from the start, not when a person is having a hard month. I cover the legal side of that decision, including what Manitoba's probation rules actually require, in the probation period post. A crew of ten notices who you keep long after they should have gone just as fast as they notice who you let go too soon.

Sources

Keep reading