Returners

Somebody who worked last peak knows the site, needs a short refresher rather than a full induction, and reaches standard in days.

Why they matter

They need a fraction of the induction and training.

They reach standard quickly, which on a ten-week peak is most of the value.

They know the physical demand and have chosen it anyway, which removes the largest cause of first-week attrition.

And they stabilise the floor, because a peak workforce with a core of returners has somebody to ask.

Why they get lost

Contact details held only in a system nobody exports before the records are purged.

No consent obtained to contact them later, so the data cannot lawfully be used.

Nobody assigned to do the approach, so August passes.

And a poor end to the previous peak, which is the reason a proportion would not come back anyway.

Keeping the list

At the end of peak, ask everybody whether they would like to be contacted about next year.

Record the consent explicitly, with a stated retention period — a data protection requirement and a simple one.

Hold the list outside the operational system, which will purge it.

And record something useful alongside the contact details: the area they worked, whether they reached standard, whether you would have them back.

The approach

August, or eight to ten weeks before the recruitment campaign opens.

A short message: we are recruiting for this peak, you worked with us last year, would you like to come back.

With the rate, the dates and the shift patterns, because they are comparing offers.

And a simple route to say yes — a reply, not a fresh application.

The returner process

Shorter, not absent.

Right to work re-verified, because circumstances change.

A refresher on safety and anything that has changed on site.

A walk of the areas, because layouts move.

And a shortened ramp expectation, which should be stated so nobody is measured against a new starter's curve.

The end of peak, which determines the list

People released abruptly with no notice do not return and tell others.

A stated end date given in advance.

A clear statement of who is being retained and on what basis.

A thank you, which costs nothing.

And the explicit invitation to return, which is the whole of the strategy and takes one conversation per person.

Retaining a few permanently

Peak reveals who is good, under pressure, over ten weeks.

Which is a better assessment than any interview, and the people you would hire are identified without a recruitment process.

Have a number of permanent vacancies held back for it, which is a scheduling decision made in September.

Most operations let those people go and recruit strangers in February.

The measure

Returner rate: peak hires who worked a previous peak.

Which should rise year on year and is the direct measure of whether the end of peak is handled.

Time to standard for returners against new starters.

Consent list size against peak headcount.

And permanent hires made from peak cohorts, which is the second return on the same effort.

The August task

One person, one afternoon: export the list, check consents, send the message.

Booked in the calendar in June, because in August everybody is on holiday and in September it is too late.

The list is the asset

Contact details, with consent, held outside the system that will purge them.

Plus one line each: area worked, reached standard, would have back.

Which turns a list of names into a recruitment plan, and it takes an afternoon at the end of peak when the information is fresh.

How it fails in practice

The list is held in a system that purges it, no consent was obtained, nobody owns the August approach, and the previous peak ended abruptly.

Every year the site recruits strangers.

The check

Ask what proportion of last peak's hires had worked a previous peak.

If nobody knows, the list does not exist.