The week lost 40 pints across fourteen shifts. A handover that narrows it down

Your weekly count says the bar lost 40 pints. Fine. The week was fourteen shifts, run by six people, with three deliveries, a line clean and a bank holiday in it. Which shift? Without shift-level records, any answer is a coin toss, and coin-toss answers are how good staff end up accused and real leaks end up ignored.

The fix is not counting the pub fourteen times a week. It's a one-page handover at changeover, ten minutes, and it earns its keep in the first month.

Ask "when" before you ask "who"

Start with an uncomfortable fact from the metering people: Vianet's insight report, from around 11,000 metered sites, found 2% of drinks poured never reach the till. Spread across a week that reads as theft. Looked at shift by shift, most of it turns out to be recording, and specifically recording that happened on the wrong side of a shift boundary.

A delivery arrives at 4pm and gets booked in whenever someone gets to it. Wastage from the lunch session gets written up at close, inside the evening shift's window. A prepaid pint is paid for on the early shift and poured on the late one, so one shift shows a sale with no stock movement and the other shows stock gone with no sale. A line clean lands its waste in whichever window it was done. None of that is loss. All of it looks like loss the moment you try to read the week at shift level, which is why the first job of a handover sheet is moving the recording to the boundary, not catching anybody.

The one-page handover

One sheet by the till covers it, and the whole thing takes about ten minutes:

Line on the sheet	What gets written	Why
Float	Counted and signed by both people	The cash version of a stocktake baseline
Till read	The X or Z read number and time	Pins sales to the window, no arguing later
Five fast lines	A count of your five biggest sellers	Splits the week's variance into short windows
Open kegs	Which kegs are on and roughly where they're up to	A keg that "suddenly" blows mid-shift stops being a mystery
Wastage so far	Anything poured away, with a reason	Recorded in the window it happened, not at close
Owed and prepaid	Paid-not-poured, tabs, anything promised	The straddling sales that fake discrepancies

The five-line count is the working heart of it, and it's the same discipline as line checks pointed at a shorter window: opening count plus deliveries minus closing count, against till sales for the same hours. Five lines, not fifty. You're not valuing the pub at changeover, you're leaving a timestamp, so that when the weekly count comes back short you can say it happened after Tuesday's handover and before Wednesday's, which is a question with an answer instead of a mood.

What it costs versus what it catches

Ten minutes at changeover, at the £12.71 living wage, is about £2.12, call it £30 a week if you're rigorous about it. Against that: one unrecorded pint per shift, at the £5.34 average, is £74.76 a week walking out of the building, near £3,900 a year, author's arithmetic. And that's the smallest version of the problem the sheet exists to catch. The more common win is quieter: the week the count comes back fine and you didn't spend an evening interrogating your own rota, because the handover sheet already showed the "missing" firkin was Tuesday's delivery booked in late.

Reading a pattern without pointing a finger

Say the five-line counts run clean all week except across one boundary, repeatedly. Before it's a staff conversation, it's a process question: what happens at that changeover? Is that when the delivery lands, when the cellar gets worked, when the shift is one person deep and the recording slips? Fix the process and re-run the counts. Most patterns die right there.

If one window keeps leaking after the recording is properly tight, you're into the territory of the unexplained loss playbook, and you'll be walking into it with evidence instead of a hunch. What you don't do is reach for the wage packet. The law caps deductions from retail workers for cash or stock shortages at 10% of gross wages in any pay day, and it needs contract terms and notice you've probably not got in place, not legal advice, read your own paperwork. More to the point, a deduction converts a process failure into a grievance, and the stock doesn't come back either way.

Handover between shifts, handover between managers

This sheet is the within-the-day version of a discipline that scales. When a relief manager takes your pub for a fortnight, the same logic says do a signed count at the door, both directions. The shift sheet protects your Tuesday. The signed count protects your holiday. Same idea, different window: never let stock change hands without a number attached.

And if your till can report by clerk or by time band, turn it on and read it monthly against the handover sheets. The voids and refunds report tells you the same story from the till's side, and the two together answer "when" precisely enough that "who" usually answers itself, and more often than not the answer is "nobody, fix the booking-in".

Common questions

How do I find out which shift is losing stock?

Narrow the window before you blame anyone. Take your five fastest-selling lines and count just those at every changeover for a week, against till sales for the same window. That splits one weekly number into a dozen short windows, and in most pubs the discrepancy turns out to sit at the boundaries, where deliveries, wastage and prepaid drinks were recorded on the wrong side of the line.

Is a full stocktake every shift realistic?

No, and you don't need one. A full count is hours; five fast lines is about ten minutes. The point of the changeover count isn't valuation, it's a timestamp. You're buying the ability to say the loss happened before Tuesday evening rather than somewhere in the week.

Can I take a till shortage out of someone's wages?

For retail workers the law caps deductions for cash or stock shortages at 10% of gross wages in any pay day, and you need the right contract terms and notice in place first. This isn't legal advice, read your own contracts. But the better question is why the shortage is there: in my experience recording gaps at changeover produce most of them, and you can't deduct your way out of a process problem.

What causes shift discrepancies that aren't real?

Timing. A delivery booked in mid-shift, wastage poured at 4pm but written up at close, a prepaid pint bought on the early shift and poured on the late one, a line clean that hit one shift's window. Each one moves stock or sales across a boundary without anything being lost, and each one goes away when recording happens at the moment, not at close.