There are three moments a day when the till stops being a tool and becomes a job: opening, with the float to count; any mid-day changeover, with its verification; and closing, with the count and the banking. None of the three is optional, none takes thirty seconds, and all three have the unpleasant habit of landing exactly while something else is happening in the store.

Yet in most shift plans, these moments do not exist. The plan says "Anna 9 to 5". It does not say "Anna counts the float at 9, so the floor belongs to Marco for ten minutes". The difference sounds like pedantry. It is not.

Money gets counted without interruptions, or it gets counted twice

Anyone who has ever closed a till knows the enemy of counting is not arithmetic, it is interruption. The last-minute customer, the phone, a colleague with a question: every interruption mid-count means starting over or, worse, getting it wrong. Till discrepancies are born this way far more often than from genuine mistakes.

The countermeasure is protected space: whoever counts, for that quarter of an hour, does nothing else and cannot be interrupted short of an emergency. Which, in a two-person store, means exactly one thing: that quarter of an hour must be planned, because the other person needs to know the floor is entirely theirs.

Double-checking, not double effort

Where volumes justify it, the good practice of the double count (the closer counts, the opener recounts) is the best protection for everyone, starting with whoever handles the cash: two signatures on a number shields people from suspicion far better than any amount of mutual trust. For it to work, though, the handoff has to sit in the natural flow of shifts: if recounting the float is the opener's first act of the day, the check exists without costing anyone extra time.

Here too, the calendar decides whether the practice lives or dies: if opening and closing always belong to the same two people, the cross-check is routine; if they rotate randomly with nobody watching, sooner or later the store gets opened by someone who has never done a recount.

Till closing time is not store closing time

A single-till classic: the till shut at 7:30pm sharp because "then there's the count", and the 7:25 customer served with anxiety or, worse, turned away. The truth is the two times must be decoupled in the plan: if the store closes at 7:30, the count is a 7:30-7:50 block belonging to whoever closes, not an activity crammed in earlier that effectively lowers the shutter ahead of time.

Those twenty minutes exist either way. Better that they exist written down, paid and calm, than stolen from the last stretch of service.

Who touched the till should be written somewhere

Last point, which becomes crucial the day something does not add up: at any moment of the day it should be possible to reconstruct who was on the till. Not for a witch hunt, but for the opposite reason: when the shift record is clear, a discrepancy gets narrowed down quickly instead of casting a shadow over the whole team for weeks.

Giving till rituals a place in the plan (counting blocks, cover during them, who opens and who closes) is a matter of a well-built calendar: Sked Solve is designed for exactly that, and the site shows how.