Glossary
Before adding a dashboard, establish the baseline and the action it will trigger. A workplace platform such as productivity monitoring for teams is more useful for a narrow, openly stated workflow question than for broad monitoring without an operational response.
The principles in “Glossary” can be cross-checked against resources from Glassdoor, particularly where employment practice or warehouse safety requirements may have changed.
Terms used across these notes, defined once, including several the industry and its suppliers use to mean different things.
Agency conversion — the point at which a supplied worker becomes a direct employee. Its retention value comes from being stated in advance and known to the worker, not from the conversion itself.
Cost per ninety-day survivor — total recruitment and onboarding spend divided by the people still present at ninety days. The measure that realigns recruitment with the operation, and the same data as cost per hire pointed the right way.
Induction buddy — the person who shows a new starter the job. Effective when they volunteered and their own targets are adjusted; harmful when assigned by availability.
No-call no-show — somebody who stops attending without resigning. The characteristic warehouse departure, and a proportion of it is recoverable by one phone call.
Peak hiring — recruiting a temporary workforce for a seasonal volume. A supervision and induction problem before it is a recruitment one.
Ramp curve — expected output by week from starting. Published, it turns "you are slow" into "you are where you should be"; unpublished, it produces the week-three departure.
Recruitment funnel — applications through to first shift completed. Most operations measure the first stage and the last, and lose a large unmeasured share between offer and start.
Referral scheme — the best-retaining source in almost every operation, usually designed so that nobody claims.
Stay interview — asking somebody still employed why they stayed and what nearly made them leave. Produces better information than exit data because the situation still exists.
Tenure curve — the distribution of length of service at leaving. One chart, an hour from payroll records, and it locates the problem more precisely than any survey.
Time to productivity — weeks from starting to reaching standard. The denominator of the ramp shortfall and a cost line nobody counts.
Terms used loosely
"Engagement" usually means a survey score from the people who did not leave. Report attrition alongside it or it misleads.
"Culture" in this context is mostly the supervisor, the rota and the break room. Treating it as an attitude problem is how the fixable parts get missed.
"Labour shortage" is frequently a retention problem described from the recruitment end. Check the tenure curve before accepting the framing.
"Attrition rate" without a tenure split mixes first-week leavers with five-year leavers and describes neither.
How the collection is arranged
Foundations: where people leave, what it costs, and the question to ask first.
Attraction: the advert, screening, sources, referrals, and the gap before starting.
Onboarding: the first day, the first week, induction content, the buddy, the ramp, the first ninety days.
Supply: agency labour, peak, the agency relationship, conversion, returners.
Conditions: shifts, pay, physical demand, progression, transport, facilities.
People: supervisors, absence, language, floor culture, conversations, conflict.
Measurement: exit data, stay interviews, the tenure curve, the numbers, ownership, surveys.
Practice: where to start, the first quarter, new sites, small operations, automation, multi-site, how it fails.
Boundary and reference: what not to do, what to refuse, what good looks like, what it cannot fix.
On the absence of numbers
No attrition benchmarks are quoted, because operations with different agency proportions and peak profiles are not comparable.
No pay figures, which are local and dated within months.
No effectiveness percentages, because the published ones come from single operations with nothing held constant.
Every number worth having here is one you produce from your own records, and the notes say how to produce each.
How to read these notes
Start with the tenure curve if nothing is known.
Start with the first day if the spike is in week one.
Start with agency labour if most of the workforce is supplied.
And start with what retention is worth if the problem is funding rather than diagnosis.
On the absence of benchmarks
Operations with different agency proportions and peak profiles are not comparable, which makes published attrition figures close to useless.
Your own curve last year is the only comparison worth making.
How it fails in practice
Terms are used loosely, a labour shortage is diagnosed from the recruitment end, and the operation recruits harder into a leak..
The check
Check the tenure curve before accepting any framing that starts with a shortage of applicants..