Key takeaways:
- The redeployment benchmarks in circulation trace back to software vendor glossaries, and the most-repeated one is more than a decade old.
- ASA’s research, on the other hand, found there are roughly six hires for every temporary spot filled each year.
- You can calculate median lifetime days from payroll records, and it needs no agreed definition to compare to your own prior period.
Search for a redeployment benchmark and you’ll get several unreliable numbers:
- Best-in-class firms redeploy above 70% of their contractors.
- The industry average is 30 to 40%.
- Anything under 25% means you’re rehiring your own bench.
The Oorwin glossary publishes both the “6% of firms” figure and a 10 to 30% range without a primary source, and Oorwin sells an ATS with a bench management module. And the origin of that 6% number is older than you may think. Recruiting automation vendor Sense published it in October 2018, citing a 2014 CareerBuilder guidebook.
The data don’t even agree with each other. Depending on the page, the industry average is 30 to 40%, or 10 to 30%, or roughly 3% for second assignments. And, oftentimes, the pages publishing this data sell software that tracks redeployment.
So what number can you trust?
ASA doesn’t report a redeployment rate. Neither does Bullhorn’s GRID report, but it does show that 85% of firms that have a redeployment plan report time to place under 20 days. That’s a correlation with placement speed, but it doesn’t tell us what share of contractors get a second assignment.
According to ASA, staffing companies employed nearly 2.2 million temporary and contract workers in an average week in 2024, and hired 12.7 million of them over the course of 2023.
That’s about six hires for every seat filled in a given year. The industry replaces the average temporary position roughly six times annually, and it pays full acquisition cost each time.
Lifetime days is the number to calculate
The reason no one agrees on a redeployment rate is that no one agrees on the denominator. Should we count assignments that ended, or contractors on the roster, or candidates placed in the period? Pick a 30-day gap or a 60-day gap between assignments and you’ll end up with very different percentages.
Median lifetime days avoids that. You take the date of a candidate’s first placement, the last day they worked, and count. You can find these in your payroll system, so there’s no definition to argue about and no window to game.
Data from 882,004 placements and 443,182 distinct candidates over 24 months, sorted by acquisition source, offers some insights. In light industrial, referred candidates posted a median of 64 lifetime days compared to 36 for the average job board candidate, a 78% difference. Healthcare came in at 180 days against 99, an 82% difference. And travel nursing was 174 against 116, a 50% difference.
That’s one dataset from one platform, but it shows that lifetime days vary enough by acquisition source to be worth measuring, and that payroll records can produce the number at scale.
Test it on your own payroll file
Pull every candidate with a first placement in a completed period, like 2024, so the cohort has had time to run. Record the first placement date and the last worked date for each. Take the median of the difference.
Then cut it three ways: by acquisition source, by recruiter, and by vertical. Those three cuts tell you where your bench is leaking and which desks are keeping people working. A single blended number isn’t as useful, which is part of why the industry average hasn’t been helpful.
Note that a candidate first placed three months ago hasn’t had the chance to accumulate lifetime days, so including recent placements will drag your median down and make any year-over-year comparison meaningless. Use a closed cohort and keep the definition steady.
The benchmark question answers itself once you’ve done it twice. Your own prior-period number is the comparison that counts, and unlike the 30 to 40%, you’ll know exactly how it was calculated.
