RetailPOS.AI
Playbooks

Stop reconciling the cash drawer by hand

Last reviewed 2026-08-13 · by the RetailPOS team

In short
If closing up means counting the drawer, typing it into a spreadsheet, and squinting at a variance with no way to explain it, the problem isn't your arithmetic — it's that the drawer has no source of truth. Real cash accountability needs four things tracked automatically: the opening float, every drop, pay-in and pay-out, the expected cash at close, and the variance against what was counted — per shift, per shop.

A cash variance you can't trace is just anxiety. A cash variance tied to a specific shift, float and movement log is a question you can actually answer. Here's how to get there.

The nightly-spreadsheet problem

Manual reconciliation breaks in predictable ways: nobody recorded the $40 taken out to pay a delivery, the opening float was “about $150,” a mid-shift cash drop wasn't written down, and now the count is $63 off with no way to reconstruct why.

So the variance either gets shrugged off (and real leaks hide inside the noise) or someone spends 20 minutes a night re-deriving numbers that the till should have kept all along.

What real drawer accountability needs

A recorded opening float — the exact cash you started with, not a guess.

Every cash event logged as it happens — drops to the safe, pay-ins, and pay-outs (the delivery driver, the window cleaner) captured at the moment, not reconstructed later.

An expected figure the system computes — not you doing the sum in your head.

A variance against the physical count, tied to that one shift and shop, so a discrepancy points somewhere instead of nowhere.

How RetailPOS does it

You open a shift with the float you put in the drawer. Through the shift, drops, pay-ins and pay-outs are recorded as cash movements as they happen.

At close, RetailPOS computes what the drawer should hold:

expected = opening float + cash tenders − cash change + pay-ins − drops − pay-outs

You count the drawer, and it shows the variance against that expected figure — flagged green, amber or red — so a clean night is obvious and a real gap stands out. The same math drives the Z-Report per shop, per day, and only one shift is open per shop at a time, so there's never ambiguity about which drawer a number belongs to.

The upshot: no spreadsheet, no head-math, and a variance you can trace to a shift and its movement log instead of a mystery you write off.

A drawer-close checklist

  • Open every shift with a counted float, not an estimate
  • Log drops, pay-ins and pay-outs the moment they happen
  • Let the system compute expected cash — don't do it by hand
  • Count, record, and read the variance at close
  • Review amber/red variances against that shift's movement log
  • Reconcile per shop, per day, from the Z-Report

Frequently asked

Why is my cash drawer always off?
Usually because cash events aren't captured as they happen — an unrecorded pay-out, a fuzzy opening float, or an unlogged drop. Track the float and every movement and most “mystery” variances disappear.
How does RetailPOS calculate expected cash?
Opening float + cash tenders − cash change + pay-ins − drops − pay-outs, per shift. You count the drawer and it shows the variance against that figure, flagged green/amber/red.
Can I reconcile per shop across multiple locations?
Yes — the Z-Report runs per shop per day, and only one shift is open per shop at a time, so each variance belongs unambiguously to one drawer.
Want the product side? See the retail pack →

Open your shop in 30 seconds.

No card. Free until your first 100 sales. No cut of your sales; keep your hardware.