How should a store keep cash sales accountable when the POS system goes down mid-shift?

Write the outage rules before the register ever goes dark

Every cash close rests on a comparison between what the drawer holds and what the system says it should hold. An outage removes the second half of that sentence. If nobody has decided in advance how sales will be captured while the terminal is dark, the store ends up with a pile of loose paper, a drawer nobody can judge, and a manager guessing at midnight. The fix is not clever, it is prepared: a one page outage procedure taped inside the office cabinet, reviewed with every closing manager, and rehearsed at least once so the first person to face it is not inventing a method under pressure. Related: How do you reconcile the till at end of day without a long headache?

The procedure should answer four questions in plain language. Who declares the outage and starts the manual process. What the store will and will not accept while the system is down, which for most operators means cash only, because card authorization and gift balance lookups depend on the same connection that just failed. Where the manual slips, the calculator and the spare pens live, in one labeled box that is not raided for other purposes. And who is authorized to reopen normal operations. Deciding this on a quiet Tuesday costs twenty minutes. Deciding it during a Friday dinner rush costs a whole close.

Keep reading: How do you run a guided cash drawer count that is right every time?, How do you track over and short by cashier to find real problems?, How do you reconcile the till at end of day without a long headache?. See how TillClosr helps you cash drawer counts and end-of-day till reconciliation.

Freeze a baseline count the moment the system fails

The single most useful action in the first five minutes is a count of the drawer as it stands. Whatever the system last recorded, that count becomes the known starting point for the outage period. Write it on the top of the manual log with the time, have a second person verify it, and both initial. From that moment the arithmetic is simple and defensible: baseline plus manual cash sales, minus manual paid-outs and change given from the drawer, minus any drops, equals what the drawer should hold when the outage ends. Related: How do you run a guided cash drawer count that is right every time?

Do the same in reverse when the system comes back. Count again, note the time, and close the outage window. Two counts and a stack of slips give you an outage that reconciles on its own terms, independent of whatever the POS reports afterward. Without those counts you are trying to explain a full shift of activity with half the record, and the variance you find at close belongs to nobody in particular. Managers sometimes skip the baseline because the store is busy and the line is growing. That is exactly the shift where the missing number will cost the most.

Make the manual slips something a close can actually reconcile to

Loose scraps of receipt paper are not a sales record. Use pre numbered duplicate slips from a carbon pad so the sequence itself proves nothing was thrown away, and keep the pad in the outage box rather than in a drawer where it gets borrowed. Each slip needs the same handful of fields every time: slip number, time, items or department, subtotal, sales tax, total, payment type and cashier initials. Tax matters more than people expect during an outage, because the register normally calculates it and now a person has to. A small laminated tax table for common price points removes most of that error. Related: Why should you keep an accurate deposit log for every bank drop made?

Keep one cashier on the drawer for the whole outage if staffing allows. Shared accountability is already the hardest problem in cash handling, and it gets worse when the audit trail is handwriting. If a shift change is unavoidable, close the outage window with a count, start a fresh one, and staple the slips to the matching count sheet. Refunds during an outage should simply wait unless a manager approves and signs the slip. A handwritten refund with no system record is the easiest possible way for cash to leave a drawer without a trace. Related: How do you track over and short by cashier to find real problems?

Rebuild the close once the system comes back

When the terminal returns, resist the urge to send everyone home. The slips still need to become sales, both because tax reporting depends on it and because the close will not make sense otherwise. Most operators ring the manual sales in as a batch, in slip order, tendered as cash, then staple the register journal to the slips. Some systems offer an offline or recovery mode that captures the transactions automatically, and if yours does, still count the physical slips against the recovered count before you trust the number. Two records that agree are worth far more than one record that arrives quickly.

Flag the close itself. An outage day should be marked in your records so that the over and short figure is never compared to a normal night without context, and so anyone reviewing the pattern later knows why the variance sits outside the usual range. Note the outage start and end times, who ran the drawer, the baseline and closing counts, and the batch total entered afterward. If a shortage appears, that packet is what turns an uncomfortable conversation into a factual one. Most outage variances turn out to be tax rounding or change given in a hurry, and the paperwork is what lets you prove it rather than argue it.