Open study
Canadian light-industrial staffing benchmarks (2026)
There are no results on this page yet. Recruiting participants. No results published yet. What is published here is the method: what we will measure, exactly how each figure is defined, how large the sample has to be before we say anything, and what we have committed not to do.
Nobody publishes this data for Canada. The numbers a Canadian agency owner finds are American, from vendors selling into a different market under different labour rules. We would rather build the real one slowly than quote theirs quickly.
Method last updated . Published by Kordis, which sells software to the agencies being measured. Judge the method accordingly, and hold us to the commitments at the bottom of this page.
What is being measured, and how each figure is defined
Definitions do most of the work in a benchmark, and a vague one is how a flattering number gets manufactured without anybody lying. These are dated and fixed.
- Fill rate
- Shifts covered by a confirmed worker who attended, divided by shifts the client ordered. Counted per shift, not per order, so a 10-worker order that fills 8 counts as 8 of 10 rather than as one miss.
- Time to fill
- Minutes between a shift becoming open and a named worker confirming it. Measured from the open event, not from when the office noticed, because the gap between those two is part of what we want to see.
- No-show rate
- Confirmed workers who did not attend, divided by confirmed workers. A worker who cancelled with notice and was replaced is not a no-show; a worker who simply did not arrive is.
- Same-day coverage
- Of the shifts that lost their worker inside 12 hours of the start time, the share that were covered before the shift started. This is the number the market cares about and the one nobody publishes.
- Office minutes per filled shift
- Coordinator time spent on contact attempts, divided by shifts filled. Self-reported for a single named week, because nobody has this instrumented and a made-up precision would be worse than an honest estimate.
How the sample is collected
- Participants are Canadian staffing agencies placing light-industrial, warehouse, or logistics workers, with 15 to 250 active workers.
- Each participant reports one consecutive four-week period, chosen by them, from the twelve months before submission. They name the period; we do not let anyone pick their best week after seeing the numbers.
- Figures come from the participant's own records, in whatever system they keep them. We do not require a particular tool, and Kordis customers are counted separately from non-customers.
- We publish nothing until at least 30 agencies have reported. Below that, the spread is wider than anything we could honestly claim from it.
- Every published figure states n, the period, and the province mix. Where a cut has fewer than 5 agencies in it, we report that it was too small rather than publishing it.
- Participants are anonymous in the published results. No agency, client, or worker is named.
What we have committed not to do
- We will not publish a number before the sample is reached, and we will not fill the gap with a figure from a US vendor's blog.
- We will not compare Kordis customers with non-customers as though it were a controlled trial. Agencies choose their tools; that is not a randomised assignment and we will not present it as one.
- We will not report a median as an average or a best case as a typical one.
- We will not quietly change a definition after seeing the data. The definitions above are dated, and any change will be shown as a change.
Take part
If you run a light-industrial, warehouse, or logistics desk anywhere in Canada, you can contribute one four-week period. You get your own figures back against the anonymised distribution before anything is published, which is the part most owners have never had: knowing whether your fill rate is good for this market or ordinary. We are starting in Ontario, British Columbia, and Alberta and will take reports from any province.
Email info@kordisai.com with the subject “Benchmark study” and we will send the one-page form. You do not have to be a Kordis customer, and taking part is not a sales conversation.
Common questions
Are there results yet?
No. This page publishes the method and the definitions before the data exists, so they cannot be retuned later to flatter the numbers. Results appear here once at least 30 Canadian agencies have reported a four-week period.
Who can take part?
Any Canadian staffing agency placing light-industrial, warehouse, or logistics workers, with roughly 15 to 250 active workers. You do not need to use Kordis, and agencies that do are counted separately so the comparison stays honest.
What do participants get?
Your own four-week figures back, set against the anonymised distribution, before anything is published. That is the part most agencies have never had: knowing whether an 82 percent fill rate is good in this market or ordinary.
Is my agency identified?
No. Published results carry no agency, client, or worker names, and any cut of the data with fewer than five agencies in it is reported as too small rather than published.