Tracking cash variance across multiple stores
To track cash variance across multiple stores, log every drawer's over/short in one shared sheet with identical columns at each store, then compare stores on absolute variance per $1,000 of cash sales rather than on net totals. Net totals hide errors, because a $20 over and a $20 short add up to zero.
One log, same columns at every store
| Date | Store | Register | Shift | Cashier | Cash sales | Expected | Counted | Over/short | Counted by | Note |
|---|---|---|---|---|---|---|---|---|---|---|
Rules: the store name is spelled the same way every time, over/short is a signed number (−2.80, not "2.80 short"), and each store uses the same fixed float per register type (see starting cash float). The per-drawer math is in cash drawer over/short.
Three numbers per store per week
| Number | How to compute | What it shows |
|---|---|---|
| Net variance | Sum of over/short with signs | Cash actually gained or lost |
| Absolute variance | Sum of over/short ignoring signs | How much counting and change error there was |
| Absolute variance per $1,000 | Absolute variance ÷ (cash sales ÷ 1,000) | Error rate, comparable between a busy and a quiet store |
Worked example: three stores, one week
| Store | Daily over/short | Cash sales | Net | Absolute | Absolute per $1,000 |
|---|---|---|---|---|---|
| A | −1.20, +0.85, −2.10, +0.40, −0.75, −1.30 | $6,200 | −$4.10 | $6.60 | $1.06 |
| B | +12.00, −11.50, +9.25, −10.00, −0.75 | $3,900 | −$1.00 | $43.50 | $11.15 |
| C | −5.00, −8.50, −0.25, −6.75, −10.00, −7.00 | $8,450 | −$37.50 | $37.50 | $4.44 |
Example math: Store B's net is 12.00 − 11.50 + 9.25 − 10.00 − 0.75 = −$1.00, the best net of the three. Its absolute variance is 12.00 + 11.50 + 9.25 + 10.00 + 0.75 = $43.50, and 43.50 ÷ 3.9 = $11.15 per $1,000 of cash sales, the worst error rate. Store C's net and absolute are both $37.50 because every day was short. Store A: 6.60 ÷ 6.2 = $1.06 per $1,000.
Ranked on net, B looks best. Ranked on error rate, B is the store with a counting or change problem, and C is the store where cash leaves in one direction every day.
Patterns and what they point to
| Pattern | Points to | Next step |
|---|---|---|
| Large overs and shorts that cancel (Store B) | Wrong change, miscounts, card/cash keyed wrong | Two-person counts for a week; match card batch daily |
| Short every day (Store C) | Missing paid-out slips, float error, or cash removed | Check the float count at open; sort the log by cashier and shift |
| Shorts only on one shift | Whoever closes or counts that shift | Switch to blind counts; rotate the counter |
| Shorts follow one cashier across stores | That cashier's drawers | Review their voids, refunds and no-sales (see signs of register theft) |
| Small variances of a few cents everywhere | Five-cent cash rounding not recorded by the register | Set one rounding rule for all stores |
On the last row: the US Treasury's Penny Production Cessation FAQ (December 23, 2025) says businesses may round, when penny change is not available, the final amount of a cash transaction to the nearest five cents, applied in a fair, consistent and transparent manner. Each rounded sale moves the drawer by at most two cents, so 50 rounded sales can move it by up to 50 × $0.02 = $1.00.
Weekly review in six steps
- Filter the log to the last 7 days.
- For each store: net, absolute, absolute per $1,000 (table above).
- Sort by absolute per $1,000; look at the top store first.
- Within that store, pivot by cashier and by shift.
- Read every note on variances above your threshold.
- Write one action per store, with a name and a date.
Shared employees need one rule: the cashier column records the person, not the home store, so a cashier who works at two locations shows up under one name in both.
For scale: the National Retail Federation's 2023 National Retail Security Survey (177 retail brands) put average shrink at 1.6% of sales in fiscal 2022, with internal and external theft making up 65% of it. Drawer logs measure the cash part only.
Assistlify records cash-drawer counts per location and lets staff be assigned to more than one location, which covers the per-store counts in the log above. You can try it in the preview workspace; more on running several stores is on the multiple locations page.
Questions
What is cash variance?
Cash variance is counted cash minus expected cash for a drawer, where expected cash is float plus cash sales minus paid-outs.
How do you compare cash variance between stores of different sizes?
Divide absolute variance by cash sales in thousands; $43.50 on $3,900 of cash sales is $11.15 per $1,000.
Why is net cash variance misleading?
Overs and shorts cancel: +$12.00 and −$11.50 net to +$0.50 while $23.50 of errors happened.
How frequently should I review cash variance?
Weekly works for the table here: seven days of drawers per store, ranked by absolute variance per $1,000.