This gets treated as a philosophy question. It isn't — it's situational, and the same employer should answer it differently for different roles in the same month. Six questions, a decision table, and the two ways operations get it wrong.
This gets framed as a philosophy question — are you the kind of operation that uses agencies, or the kind that hires its own people. It isn't. It is a situational question, and the same employer should answer it differently for different roles in the same month.
Six questions settle it. Answer them honestly and the decision usually makes itself.
Start here, because it decides more than anything else.
If you know with reasonable confidence that this work exists twelve months from now, you are looking at a permanent role and you should be building toward your own payroll. Renting a permanent job indefinitely costs more over a year and, more importantly, gives the person no reason to invest in you.
If the volume is a peak, a contract you might not renew, a leave to cover, or a new line you are still proving out — that is uncertainty, and uncertainty is what temporary staffing is actually for. The value is not cheap labour. It is not having to commit before you know.
The honest middle case: you think it's permanent but you're not sure. That is what temp-to-hire exists for, and it is the most under-used arrangement in the industry.
A direct hire for a floor role realistically takes two to five weeks end to end: writing and posting, waiting for applications, screening, interviewing, references, offer, and any notice period. That is not inefficiency; it is the process working.
If the seat can be empty for four weeks without real cost, hire directly. If it can't — because you're covering a shipment window, a shutdown, a harvest, or a shift that's already running short — then speed has a price, and the question becomes whether the markup is less than four weeks of your vacancy cost per day. Often it is, by a wide margin.
For most general labour, warehouse and light production roles, the answer is yes: the physical work and the site-specific process are learned in days, and what matters is reliability, attitude and physical suitability. Those are exactly the attributes an interview is bad at predicting and two weeks on the floor is excellent at revealing. Strong case for starting someone temporary.
Where the skill is not teachable in a week — a licence class, a specific certification, a narrow machine, a system you can't train on quickly — the calculus changes. Those people are scarce, they know it, and they are usually looking for permanent positions. A temporary posting for a scarce certified role will get you a thin field. Hire directly and pay properly.
This is the question nobody asks out loud. Direct hiring means you carry the recruiting hours, the payroll setup, the records, the workers' compensation administration, and the termination process if it doesn't work.
For a company with an HR function, that's a Tuesday. For a 25-person operation where the plant manager does the hiring between shifts, those hours come directly out of running the plant — and they are the reason so many small operations hire the first available warm body and then live with the consequences for six months.
If your screening is honestly "whoever showed up," an agency isn't a luxury. It's the screening you don't currently have.
If you have a clear, written standard for the role — output, quality, attendance, a 30-day checkpoint — you can hire directly with confidence, because you'll know within a month and you'll act.
If you don't, direct hiring tends to produce a specific failure: a marginal employee who never quite gets managed out, absorbs supervisor attention for a year, and quietly sets the floor's standard. A trial period with a defined end date forces the decision that otherwise doesn't get made.
Be concrete. With a direct hire you own the termination, the administration, the possible dispute, the team's reaction, and then the whole search again from zero — with the seat empty throughout. With an agency placement that goes wrong, you make a phone call, and a serious agency replaces the person under whatever its written guarantee says.
That difference in unwinding cost is a large part of what the markup buys, and it is worth more the smaller your operation is and the tighter your labour market is.
| Situation | Lean | Why |
|---|---|---|
| Seasonal peak, harvest, shutdown, known end date | Agency | Defined duration, no permanent commitment, speed matters |
| Maternity, medical or extended leave cover | Agency | Fixed term by definition; the permanent role already has an owner |
| New shift or new line you're still proving out | Agency, then convert | You don't yet know whether the volume holds |
| Someone resigned and the seat costs you daily | Agency now, hire in parallel | Cover the gap while running a proper search — don't rush the permanent decision |
| Stable, year-round, well-defined position | Direct hire | Cheaper over twelve months and the worker has a reason to stay |
| Scarce certification or licence class | Direct hire | Thin candidate pool that wants permanence and pay |
| Supervisory or team-lead role | Direct hire | Authority needs standing; temporary status undercuts it |
| Strong referral pipeline, fills in days | Direct hire | Low vacancy cost, low failure rate — use what works |
| Same position failing three or four times a year | Neither — fix the role | A sourcing solution to a retention problem just repeats the bill |
The best-run floors we work on don't choose. They run a core-and-flex model, and it looks like this:
The mistake at each extreme is symmetrical. All-permanent operations end up carrying idle labour through slow periods, then panic-hiring in peaks. All-temporary operations never build institutional knowledge, so every peak is staffed by people who have to be taught the basics again.
The question is not agency or direct hire. It's which portion of your labour is genuinely stable — and being honest about how small that portion is.
Using an agency as a permanent staffing strategy. If someone has been on our payroll doing your core job for a year, that is a decision being avoided. It costs you more than converting, and the worker knows they're renting a chair. Convert them.
Direct-hiring into genuine uncertainty. Hiring permanently for a peak you hope will hold produces layoffs three months later — expensive, damaging to your reputation in a small local labour market, and entirely avoidable. In Halifax or Moncton, where the pool is finite and word travels, this is much more costly than it looks on a spreadsheet.
Related reading: how temporary staffing actually works for the mechanism and the questions to ask an agency, and the seasonal hiring calendar if your answer to question one is "it depends on the month."
When the volume is uncertain, the seat costs you money every day it's empty, the skill is teachable in a week, or you don't have the admin capacity to screen and onboard properly. Seasonal peaks, leave coverage, a new shift you're proving out, and covering a sudden resignation while you run a proper search are the four clearest cases.
When the role is stable and year-round, when it needs a scarce certification or licence class, when it carries supervisory authority, or when you already have a referral pipeline that fills roles in days. In all four your vacancy cost and failure risk are low and permanence is what the candidate wants.
A permanent core of directly hired people who hold the knowledge and set the standard, plus a flexible temporary layer sized to the peak above that baseline — so the core is never laid off in a slow quarter and never overwhelmed in a busy one. Good temporary workers then become the pre-screened pipeline for the next permanent opening.
Usually yes — it means a decision is being avoided. Over twelve months it costs more than converting, and the worker knows they're renting the chair, which shows up as turnover. If the role has proved permanent, convert the person.
Realistically two to five weeks end to end — writing and posting, waiting for applications, screening, interviews, references, offer and any notice period. If your seat can be empty that long without real cost, hire directly. If it can't, that gap is what you're actually buying from an agency.
Ten minutes on the role and the volume, and we'll give you a straight recommendation — including the cases where you shouldn't be paying an agency at all.