How would you know if your redeployment rate was good?

There isn't a definitive industry benchmark for redeployment. If one agency reports redeploying 60% of its contractors and another reports 20%, you can't tell which is performing better without knowing how each of them counted.

A published figure still tells you something. It just doesn't give you a reliable basis for comparison.

That matters because redeployment is one of the cheapest revenue channels a temp or contract agency has access to. The worker is already sourced, already screened, already cleared for the work they have been doing, and already proven in front of a client. This covers why the published numbers don't line up, what should happen in the four weeks before an assignment ends, and how to build a number of your own that means something.

Why is there no useful redeployment benchmark?

Because the measurement window changes the answer, and almost nobody publishes theirs.

Redeployment rate is usually described as the share of contractors who finish an assignment and start another one with you. The definition sounds tight. The measurement isn't. An agency counting across a sixty-day window includes contractors who spent two months working somewhere else before coming back. An agency counting across fourteen days doesn't. Same performance, very different number, and neither is wrong.

The denominator varies too. Contractors who retired, went permanent with the client, or left the country can be counted or excluded. Whether extensions count as redeployments is a judgement call agencies answer differently.

None of that makes the published figures dishonest. It makes them incomparable.

The asset agencies say they value most

In our FARR Report 2026, based on responses from 132 agency leaders across the UK and Ireland, the existing candidate database was rated positive or excellent ROI by 80% of agencies, ahead of job boards at 72%.

Agency leaders already believe the database is their strongest asset. The question is which part of it.

A contractor currently on assignment is the most qualified record you hold. Someone has sourced them, screened them, cleared them for the work they are doing, and watched them perform in front of a client who kept them.

Some of that clearance travels and some doesn't. A carded, inducted worker on one construction site isn't automatically ready for the next, because clients run different checks, different schemes and their own inductions. What does travel is everything you learned about whether they turn up, work well and get asked back. That's the part you can't buy.

And in most agencies it's the record least likely to have a process attached, because the assignment is going well and nothing appears to need attention.

What should happen before an assignment ends

An assignment ending is a scheduled event. The four weeks before it are three decisions, not three diary reminders.

Four weeks out: the client's intention. Does the requirement continue, and in what shape? This is a commercial conversation about the client's next quarter, not an administrative check on a date. It also surfaces demand you didn't know about, because a client deciding whether to extend one worker is usually thinking about the wider team.

Two weeks out: the worker's intention. What do they want next, when are they available, and what would they accept? A contractor who has heard nothing from you by this point has usually started answering other people's calls.

One week out: confirm and act. Whichever outcome the first two conversations point to, the work of setting it up starts now rather than on the last day.

Almost every agency leader would recognise that sequence as correct. The reason it doesn't happen isn't disagreement.

Four outcomes, not two

Most agencies treat the end of an assignment as a binary: it extends or it doesn't. There are four outcomes, and three of them are good.

The assignment extends with the same client. The worker is redeployed to a different client. The worker returns to your available pool with a known date and confirmed preferences. Or the worker is lost, to another agency or out of the market.

Not every outcome is available to you, though. Client terms often restrict moving a worker to a competitor or to another of that client's suppliers, transfer and temp-to-perm fees can make a move uneconomic, and PSL or MSP arrangements sometimes remove the option altogether. Knowing which of your accounts allow redeployment is a prerequisite to the conversation, not something to discover halfway through it.

The second outcome also has to be sold. The receiving client hasn't met the worker, and in contract recruitment that means a submission, sometimes an interview and always a rate conversation. The advantage isn't that you skip the process. It's that you go into it able to speak to someone's reliability from evidence rather than from a reference.

One wrinkle undermines the metric everyone quotes. If extensions count as redeployments, your number goes up. If they don't, an agency that extends brilliantly will look poor on redeployment rate, because its best-managed contractors never come off assignment to enter the denominator. The commercially strongest outcome is invisible to the measure. That's another reason two agencies' figures can't be compared, and a good reason to track extensions separately in your own reporting.

Whether they take your call was decided weeks ago

The four-week conversation only works if the worker wants to hear from you, and that's settled by the assignment they've just finished rather than by how well you time the call.

Pay is the largest factor. A worker whose pay was wrong once will usually give you the benefit of the doubt. A worker whose pay was wrong twice has already decided, and no amount of process at the four-week mark will recover it.

Shift accuracy matters nearly as much. If the hours, the location or the rate turned out different from what they were told at booking, the trust required to accept an unseen next assignment isn't there.

The third factor is whether anyone spoke to them at all. A worker who heard from you on day one and then not again until you needed something has been treated as a resource rather than a relationship, and they know it.

Redeployment is often filed under recruitment. Most of what determines it happens in operations.

Why the end date isn't the trigger

Every agency records assignment end dates. Very few act on them consistently. The gap between those two facts is the whole problem.

An end date is passive. It records when something finishes. It doesn't instruct anyone, it isn't owned by a named person, and it surfaces only if somebody opens the view it lives in. A field in a record is not a workflow, and treating it as one is why the conversation tends to happen on the last Friday rather than four weeks out.

In Firefish, contractors approaching the end of a contract are tracked alongside renewal and end dates, and their availability and compliance status feed into matching.

Who owns each conversation?

The two conversations usually belong to two different people. The client intention conversation sits with whoever owns that account. The worker intention conversation sits with the recruiter who placed them, because that's the relationship the contractor recognises. Both are correct, and neither person can complete the sequence alone. In a smaller agency where one person holds both, they compete with live vacancies in the same week, and both lose.

In Firefish, candidate availability is captured through the candidate portal as well as by recruiters, and bookings stay connected to the candidate, client and job records, so both conversations draw on the same information.

The clock is different in every sector

The sequence holds everywhere. The timing doesn't.

In construction, a four-week countdown may be most of the assignment, so the conversation has to start almost as soon as the worker is on site. Engineering and energy sit at the other end, where six- and twelve-month contracts give real runway, and four weeks is enough time to place someone properly. That's exactly why the window gets taken for granted and the conversation slides.

Logistics and warehousing are the hard case. Rolling weekly cover means the countdown never starts, because there's no end date to react to. That sounds easier and isn't, since nothing prompts the conversation at all. The substitute is a cadence rather than a trigger: a standing weekly check on availability, preferences and next available date for everyone currently working, so the pool stays current whether or not anything is ending.

How to measure your own

The point isn't that measurement is impossible. It's that the number has to be yours.

Fix the window, and never move it. Fourteen days, thirty, sixty. Any of them is defensible. None is comparable to another. Changing your own window between quarters does to you exactly what the industry has done to everyone else.

Decide whether extensions count. We’d argue they shouldn't, because an extension with the same client is a continuation rather than a redeployment, and counting both together measures two things at once. Track extensions as their own number.

Decide what comes out of the denominator. Workers who retired, went permanent with the client, or left the market. Whatever you exclude, exclude it every quarter.

Count contractors coming off assignment, not contractors on the database. The denominator is people who finished something in the period.

Track the margin alongside the rate. The count won't tell you what the redeployments were worth. A client taking on a worker they already know often expects a discount for the reduced risk, and a worker moving between assignments frequently wants more than they were on. A rate of 40% at flat margin is a different result from 40% achieved by conceding a point on every move.

Your first quarter gives you a number with nothing to compare it to. That's fine. This is a habit that starts paying from the second measurement, and a number that means something in three months is worth more than a benchmark that means nothing today.

Questions worth asking

• Which assignments end in the next 30 days, and who has spoken to those clients?

• When a contractor becomes available, how many people in the business find out, and how quickly?

• Who owns the extension conversation on your largest account, and is that written down anywhere?

• If your busiest recruiter had a difficult week, which of these conversations would be the first to be skipped?

Redeployment doesn't require new candidates, new clients or new spend. It requires somebody to have the right conversation four weeks before the date everyone can already see.