Anyone who has worked retail knows “I’ll close the till” sounds harmless until it hits the 6:30 queue. If the plan is “whoever can”, it’s often whoever is already tired or alone on the floor, and customers feel rush even when nobody wants it.
It’s not always about two extra people: it’s about not stacking the count in the same quarter-hour as the peak without telling anyone. A small shift move, a big difference.
What the team needs
To know what margin exists to start the procedure without improvising excuses to customers. If company rules impose windows, those windows belong in the schedule, not only in the lead’s head.
A common mistake
Treating the count as an “admin task” separate from service. It actually removes service for a slice of time: that belongs in the plan.
Till work and shifts get along when they stop competing for the same clock.
The count as a service blackout
A till count is not parallel work to selling—it pulls hands and attention off the floor. Block it on the rota like a hard appointment or a courier slot. If your POS only allows certain windows, those windows are shop constraints just like the 6:30 rush.
Who counts, who talks to the queue
With two people, split roles: one on the count, one greeting and saying “we’re closing the till, one moment.” Silent concentration reads as rude hurry. With a solo shift, your policy must own that trade-off explicitly—and the rota must accept the queue cost.
Dual control and repeating variances
Where policy demands it, the opener should see yesterday’s close was signed off. Patterns of variance belong in a shared column so you choose extra minutes, a second checker, or POS coaching—not just “be more careful.”
Move the count away from the spike
Look at three weeks of traffic: if the same weekday and hour always queues, do not schedule the full count then. Starting twenty minutes earlier or borrowing someone for fifteen minutes costs less than a complaint or a miscount.
Four fields to close the loop on differences
Date, expected/actual, hypothesis, action. Close within forty-eight hours so the next shift does not inherit mystery or double fixes.
One monthly question: how long does it really take?
Compare real count duration to the procedure sheet. If reality is double, the rota is lying—update time allowed and cover instead of demanding impossible speed.
Tell shoppers a time, not just “we’re closed”
“We’re closing the till—about five minutes” lands softer than silence and tapping. Whoever faces the queue needs a realistic range from the team, not made-up promises. Clear waits feel shorter than unexplained concentration on keys.
Mid-shift counts need the same respect
If policy forces a count between lunch and evening rush, block it like a meeting. Sales will dip for that slice—plan for it instead of pretending the floor is unchanged.
HQ windows are rota inputs, not surprises
If finance mandates upload times, publish them beside the shop rota as hard constraints. Better to shift selling cover or ask for a one-off exception than miss the window every week. Resolve “numbers vs queue” in the schedule, not over a headset at 21:55.
Train the customer-facing line, not only the count
Role-play the thirty-second explanation shoppers hear while the till is blind. If colleagues mumble or contradict each other, the queue stress doubles. One agreed phrase removes side-eye between team members mid-count.
Sked Solve
Sked Solve helps you see who is on and when, so “technical” moments don’t arrive as surprises. Start your free trial.