A bad hire costs you more than the wages you paid them. By the time you add the wages during a failed trial, the notice you owe under Manitoba law, and the cost of running the search again, one wrong pick can cost a 10-person Manitoba crew several thousand dollars, most of it before you have even reposted the job.
What does the research actually say about the cost of a bad hire?
Most of what gets published on this is American, and I am not going to pretend otherwise or round it into a Canadian number it is not. Here is what I could actually verify on the source's own page, with the year and the currency attached.
A Center for American Progress review of 30 studies found that replacing an employee typically costs about one-fifth of their annual salary, and that figure mostly counts direct costs (Center for American Progress, 2012, USD). Gallup puts full replacement cost even higher, at one-half to two times the departing employee's annual salary (Gallup, 2019, USD). CareerBuilder's own December 2017 survey, run with Harris Poll, found companies lost an average of $14,900 on every bad hire, and 74 percent of employers said they had hired the wrong person at least once (CareerBuilder, 2017, USD).
The one Canadian figure I could verify on this is about time, not dollars. Robert Half Canada surveyed more than 270 CFOs and found they spend 21 percent of their working hours, on average, coaching or managing an employee who is not performing. That is just over eight hours out of a 40-hour week (Robert Half Canada, 2018). Robert Half Canada asked more than 1,500 Canadian hiring managers again in 2025: 24 percent admitted a hiring mistake in the past two years, it took them four weeks on average to recognize it, and it cost more than 15 hours of productivity a week while it lasted (Robert Half Canada, June 2025).
What I can verify for Manitoba directly is how tight the labour market already is, which is what makes replacing someone expensive in time as well as money. Across Canada, the job vacancy rate in the first quarter of 2026 was 2.8 percent (Statistics Canada, The Daily, June 16, 2026). Manitoba runs in the same tight range, and a trades shop that loses a hire re-enters that market from the back of the line. A wrong hire does not just cost you what you paid them. It costs you a slot in a market where the next candidate is not sitting around waiting.
What does this look like for a 10-person crew in Manitoba?
Say you run a 10-person Manitoba outfit and you hire a general construction labourer. Manitoba's median wage for that role is $22.00 an hour (Job Bank Canada, updated November 19, 2025). If your seat pays more, a carpenter's median wage in Manitoba is $27.00 an hour and an administrative assistant's is $24.32 an hour on the same Job Bank data, so scale the math below to your own number.
Six weeks in, it is clear this is not working out. He has been on the job less than a year, so under Manitoba's Employment Standards Code he is entitled to one week's notice, or pay in lieu (Manitoba Employment Standards). Here is what that hire actually cost you, before you have even reposted the job.
| Cost | Amount | How it is calculated | Source |
|---|---|---|---|
| Wages paid during the 6-week trial | $5,280 CAD | 6 weeks × 40 hours × $22.00/hour | Job Bank Canada, 2025 |
| Wage in lieu of notice | $880 CAD | 1 week × 40 hours × $22.00/hour (under 1 year of service) | Manitoba Employment Standards |
| Estimated lost output during the trial | $9,152 CAD | 20% of this role's annualized wage ($22.00 × 2,080 hours = $45,760/year), using the CAP share as a US benchmark applied to a Manitoba wage | Center for American Progress, 2012 |
Add up only the Manitoba dollars you actually paid out, the wages and the notice, and you are at $6,160, gone, for a hire who never got past week six. I am not adding the Center for American Progress figure on top of that number, because it is a US percentage applied to a Manitoba wage, and stacking it into one grand total would be exactly the kind of false precision I am trying to avoid in this post. What it tells you honestly is that the true cost of rerunning the search and absorbing the lost output sits on top of that $6,160, and it is not small. On a 10-person crew, this one hire was ten percent of your whole team, for six weeks of work you now have to redo.
Why does a small crew feel this harder than a big one?
None of this needs a new statistic to explain, just arithmetic. On a 10-person crew, one bad hire is 10 percent of your headcount. On a 200-person company, the same bad hire is half a percent. The dollar cost of the wages, the notice, and the redone search does not shrink because your crew is small. It is the same $6,160 whether you have 10 people or 100. But the weight of it, the missed days on the job, the work your other nine people absorbed while you were one person short, and the share of your total payroll it represents, lands heaviest exactly where you have the least room to carry it.
There is also no bench. A 10-person crew usually does not have a second administrative assistant or a spare labourer who can quietly cover a gap while you look again. The work either gets done by people already stretched, or it does not get done, and the customer notices.
What actually prevents a bad hire?
I am not going to hand you a number here, because I do not have a sourced one for how much a good process saves you. What I have is what actually changes the odds, from running this for other Manitoba owners.
Write the seat down before you post it. What does success in this role look like in 90 days? What does this person need to own without being told? Most bad hires trace back to a job posting nobody actually thought through.
Screen for character before you screen for skill. You can teach someone to run a saw or answer a phone. You cannot teach someone to show up, take ownership, and be honest when something goes wrong. An assessment that surfaces how someone is actually wired, before you have built a relationship with them and started making excuses for red flags, catches more of this than a resume ever will.
Run a real interview, not a chat. A structured set of questions, asked the same way of every candidate, beats a friendly conversation almost every time, because a friendly conversation tells you whether you liked them, not whether they can do the job.
None of this is exotic. It is the difference between hiring on purpose and hiring because someone showed up and you were desperate.
How can I help?
If you want a job posted and screened while you run the interviews yourself, that is hiring with me. From $500 a month, plus $1,000 a role: a job posting written and placed, every applicant read and screened, phone screens with the real people, candidates graded against the role's scorecard, DISC assessments for the top candidates, and a list of candidates every two weeks while the role is posted.
Frequently asked questions
How much does a bad hire actually cost?
There is no single verified number, and anyone who hands you one flat figure is rounding off more than they are telling you. What is verified: a Center for American Progress review of 30 studies found that replacing an employee typically costs about one-fifth of their annual salary, mostly direct costs (Center for American Progress, 2012, USD), and Gallup puts the full cost at one-half to two times salary (Gallup, 2019, USD). CareerBuilder's 2017 survey put the average cost of a bad hire specifically at $14,900 USD. For a Manitoba example built from real Job Bank wages and Manitoba Employment Standards notice rules, see the worked example in this post.
Is any of this Canadian data, or is it all American research?
Most of the cost-of-a-bad-hire research is American, and every US figure in this post is labelled USD so you are not mistaking it for a Canadian number. The Canadian-specific data I could verify is Robert Half Canada's 2018 and 2025 surveys on time lost managing an underperforming employee, and Statistics Canada's Job Vacancy and Wage Survey, which shows how tight the Manitoba labour market actually is.
What do I legally owe someone I let go during a trial period in Manitoba?
Under Manitoba's Employment Standards Code, an employee with at least 30 days but less than one year of service is entitled to one week's notice or pay in lieu. That climbs with tenure: two weeks at one to three years, four weeks at three to five years, six weeks at five to ten years, and eight weeks past ten years. You can have them work the notice, pay it out, or split it.
What is the fastest way to stop making this mistake?
Write a real job description before you post the role, and put every candidate through the same structured screening and assessment instead of a friendly chat. I can take the posting and screening off your plate.
Which of your services fits a 10-person company?
Most 10-person owners start by filling one seat. Hiring starts from $500 a month, plus $1,000 a role. The Hiring page has the details.
Sources
- Center for American Progress: Heather Boushey and Sarah Jane Glynn, "There Are Significant Business Costs to Replacing Employees," November 16, 2012. americanprogress.org
- Gallup: "This Fixable Problem Costs U.S. Businesses $1 Trillion," March 13, 2019. gallup.com
- CareerBuilder / Harris Poll: "Nearly Three in Four Employers Affected by a Bad Hire, According to a Recent CareerBuilder Survey," December 7, 2017. prnewswire.com
- Statistics Canada: Job Vacancy and Wage Survey, Table 14-10-0441-01, first quarter 2026, released June 16, 2026. statcan.gc.ca
- Job Bank Canada: "General Construction Labourer in Manitoba | Wages," updated November 19, 2025. jobbank.gc.ca
- Manitoba Employment Standards: "Termination of Employment," Government of Manitoba. gov.mb.ca
- Robert Half Canada: "The High Price Of A Low Performer," May 15, 2018. newswire.ca