Cash, UPI and card reconciliation, with a reason on every variance.
One row per payment mode — cash, UPI, card, and a credit or other row for what you owe customers rather than hold. Expected from the day’s sales against what you counted, the variance between them, and a written reason for every difference, plus where each figure was checked. Then a close-out block with the opening float, the bank drop, and two signatures. Copy it as CSV.
The template, filled with an example
The example is a fictional single-counter trading day with four payment modes. The variance column is counted minus expected, computed rather than typed, so you can check every row.
Illustrative example only. The business, GSTIN, location, people, and all rupee figures on this page are invented. No shrinkage rate, loss percentage, or industry figure is claimed anywhere — the example exists to show the structure, and the ₹50 and ₹500 differences are chosen to be small and explainable.
Header block
- Business
- Sample Retail Private Limited (example)
- GSTIN
- 29AAAAA0000A1Z5
- Location
- Sample Store, Bengaluru 560038, Karnataka (example)
- Trading date
- 04-10-2026
- Counter
- Counter 1 of 1 (example)
- Counted by
- Sample Counter Staff (example)
- Checked by
- Sample Store Manager (example)
- Counted at
- 21:15, 04-10-2026, after the counter was closed
Expected against counted, by payment mode
| Payment mode | Expected from the day's sales (INR) | Counted (INR) | Variance (INR) | Reason for the variance | Where it was checked |
|---|---|---|---|---|---|
| Cash | 48,200.00 | 48,150.00 | -50.00 | Short by ₹50. One note left in the drawer was found after counting; recount confirmed ₹48,150. | Drawer counted twice, in the presence of two people |
| UPI | 63,450.00 | 63,450.00 | 0.00 | No variance. Matched line by line against the UPI collection report for the day. | UPI report 04-10-2026, all entries timestamped before 21:00 |
| Card | 28,190.00 | 28,690.00 | +500.00 | Over by ₹500. One card payment of ₹500 was recorded against the wrong batch; corrected against the acquirer report. | Acquirer report 04-10-2026, settled D+1 |
| Credit / other (receivable, returned card, adjustment) | 6,300.00 | 6,300.00 | 0.00 | No variance. These are amounts owed to the customer or payable by them, not money in the drawer. | Kept out of the cash total deliberately — see the note below |
| Total (all modes) | 1,46,140.00 | 1,46,590.00 | +450.00 | Net variance is a sum of two errors pointing in opposite directions and is not a result. Read the rows. | — |
Across cash, UPI, and card the day expected 1,39,840.00 and counted 1,40,290.00. Including the credit and other row the totals are 1,46,140.00 expected against 1,46,590.00 counted, a net 450.00 — which is 2 errors pointing in opposite directions. A small net number is not a clean day.
The credit and other row is deliberately not added into money in hand. It is expected value you will be asked about later, and it belongs on the sheet so it is visible and labelled — not so it can be counted as cash. If your business has no such row, delete it and use the freed column for a mode that matters to you.
Close-out
- Opening cash in the drawer
- ₹5,000.00 (counted at open, noted in the shift log)
- Expected cash from the day's sales
- ₹1,39,840.00 across cash, UPI, and card
- Counted across cash, UPI, and card
- ₹1,40,290.00
- Net money variance
- ₹450.00 — two opposite errors, so read the rows, not the total
- Variances left unexplained at close
- 0 of 2 — both were identified and corrected the same evening
- Variances carried to tomorrow
- 0
- Closing cash in the drawer
- ₹23,150.00 — opening float ₹5,000.00 plus counted cash ₹48,150.00, less the ₹30,000.00 bank drop
- Bank drop made
- ₹30,000.00 deposited, receipt noted, deposit slip filed with this sheet
- Discounts given beyond threshold
- 2 discounts, both approved by the store manager and logged in the discount column
- Signed off by
- Sample Store Manager (example), 21:40, 04-10-2026
This worksheet is an operational record, not a tax document. It does not carry GSTIN details, HSN or SAC codes, or the CGST and SGST or IGST treatment of the invoices behind these sales — those live on the invoices themselves. Confirm your own tax obligations and reporting with your chartered accountant.
Cash / UPI / card reconciliation worksheet (CSV — illustrative example data) Field,Value Business,Sample Retail Private Limited (example) GSTIN,29AAAAA0000A1Z5 Location,"Sample Store, Bengaluru 560038, Karnataka (example)" Trading date,04-10-2026 Counter,Counter 1 of 1 (example) Counted by,Sample Counter Staff (example) Checked by,Sample Store Manager (example) Counted at,"21:15, 04-10-2026, after the counter was closed" Payment mode,Expected from the day's sales (INR),Counted (INR),Variance (INR),Reason for the variance,Where it was checked Cash,"48,200.00","48,150.00",-50.00,"Short by ₹50. One note left in the drawer was found after counting; recount confirmed ₹48,150.","Drawer counted twice, in the presence of two people" UPI,"63,450.00","63,450.00",0.00,No variance. Matched line by line against the UPI collection report for the day.,"UPI report 04-10-2026, all entries timestamped before 21:00" Card,"28,190.00","28,690.00",+500.00,Over by ₹500. One card payment of ₹500 was recorded against the wrong batch; corrected against the acquirer report.,"Acquirer report 04-10-2026, settled D+1" "Credit / other (receivable, returned card, adjustment)","6,300.00","6,300.00",0.00,"No variance. These are amounts owed to the customer or payable by them, not money in the drawer.",Kept out of the cash total deliberately — see the note below Total (all modes),"1,46,140.00","1,46,590.00",+450.00,Net variance is a sum of two errors pointing in opposite directions and is not a result. Read the rows.,— Close-out,Value Opening cash in the drawer,"₹5,000.00 (counted at open, noted in the shift log)" Expected cash from the day's sales,"₹1,39,840.00 across cash, UPI, and card" "Counted across cash, UPI, and card","₹1,40,290.00" Net money variance,"₹450.00 — two opposite errors, so read the rows, not the total" Variances left unexplained at close,0 of 2 — both were identified and corrected the same evening Variances carried to tomorrow,0 Closing cash in the drawer,"₹23,150.00 — opening float ₹5,000.00 plus counted cash ₹48,150.00, less the ₹30,000.00 bank drop" Bank drop made,"₹30,000.00 deposited, receipt noted, deposit slip filed with this sheet" Discounts given beyond threshold,"2 discounts, both approved by the store manager and logged in the discount column" Signed off by,"Sample Store Manager (example), 21:40, 04-10-2026"
The reason column is the whole document. A variance without one is a number that will be rediscovered next month as a pattern, and the pattern is much harder to trace back than the note would have been. The example’s two variances are both resolved in the same evening — ₹50 found in the drawer, ₹500 recorded against the wrong card batch — which is what closing the day quickly enough to still act on it looks like.
How to fill it in
Five steps, done at the counter rather than at the end of the month.
Write the expected figure from the sales record, not from memory
Before anybody opens the drawer, the day's sales for each mode come off the billing system or the sales register. The example's expected column comes from the day's transactions, which is the only version of “expected” that can be argued with later.
Count cash with a second person present
Count the drawer twice, note the opening float, and record who was there. The example notes the opening cash in the close-out block because a variance cannot be understood without knowing what was in the drawer before the first sale.
Verify UPI and card against their own reports
UPI against the collection report for the day, card against the acquirer report — noting that card settlement usually lands the next day. Record which report you used, as the example does, so a variance can be traced to a source rather than argued about.
Give every variance a reason, in writing, at the time
Not at the end of the month. The example's ₹500 card overage was found by matching against the acquirer report that evening, and a mis-tagged payment found the same evening is a two-minute correction. Write “unexplained” honestly if that is the truth, and carry it forward.
Sign it off with a name and a time
Two people, one signature each, and a timestamp. A reconciliation with no name against it has no owner the next time the same variance appears, and the second occurrence is the one that starts being expensive.
When to use this, and when to stop
A reconciliation is only worth doing if it happens while the day is still recoverable.
When this template is the right tool
- You run a counter or a shop floor and one or two people close the day.
- You take payment in several modes and want to know which one the difference is hiding in.
- You are still learning whether your variance is a counting problem, a tagging problem, or an adjustment problem.
- You want a written record that two people agreed the drawer was counted, rather than an assumption that they did.
When you have outgrown it
- Expected figures are reconstructed at the end of the month because nobody has the day's sales per mode to hand.
- UPI and card are never verified against their own reports, so a settlement difference is discovered at month end.
- The drawer is counted by one person at the end of a long shift, with no second set of eyes and no note of the opening float.
- Variances are absorbed into the next day's float and nobody can say when the drift started.
The same process in NoxOrigin
The worksheet asks a person to reconstruct a comparison at the end of a long day. In NoxOrigin the comparison is produced from the transactions: shifts and the payment mix come off the day’s own sales, expected cash is what the billing records say it should be, and variance and exceptions appear as items to explain rather than as a reconstruction from memory. Payment modes are recorded at the point of sale — cash, UPI, card, and anything taken on credit — so the split is a fact rather than an estimate made at closing time, and the person who counted can be checked against the person who billed. Nothing here replaces your accountant or your accounting package; the point is that the figures reaching them are already a comparison instead of a reconstruction.
- Day-End Business Reporting in NoxOriginShifts, payment mix, expected cash, counted cash, variance, and exceptions produced from the transactions, so closing the day is a comparison rather than a reconstruction.
- Day-end reconciliation calculatorThe calculator counterpart to this worksheet: work out expected cash, counted cash, and variance on your own figures.
- Quotes, Billing & Finance in NoxOriginWhere the day's invoices and the recorded payments behind these figures come from, with the tax treatment set once per invoice.
- GST billing softwareThe invoice the expected figure is built from, with HSN or SAC, CGST and SGST or IGST, and place of supply handled on the document.
- Day-end closing sheetThe wider close-of-day document this mode-by-mode comparison sits inside, with discounts and exceptions to review.
- Payment collection trackerFor the amounts on the credit and other row — what a customer still owes, how old it is, and the next action against it.
Cash, UPI and card reconciliation questions
How is this different from the day-end closing sheet?
The closing sheet is the wider routine: the shift, the payment mix, discounts, and the exceptions to review at the end of a trading day. This worksheet is narrower on purpose — one row per payment mode, expected against counted, with a reason attached to every variance and somewhere recorded that the check actually happened. If you run a counter, use the closing sheet as the shift document and this as the mode-by-mode comparison inside it.
Why keep credit and other on the same sheet if it is not money in the drawer?
Because it is still expected value that a customer will come back about, and putting it on the same sheet stops it being argued about separately later. The example keeps it as its own row, states clearly that it is excluded from the cash total, and never adds it into money in hand. What matters is that it is visible and labelled, not that it is summed in.
What should I do about a variance I cannot explain?
Write it down as unexplained and carry it forward, rather than closing the day to a clean total. An unexplained variance of ₹200 that gets carried and watched is information; one that is quietly absorbed into the next day's float is a small error that gets repeated until it is large. The example resolves both of its variances in the same evening, which is the version worth aiming for rather than the one to assume.
When do I need my chartered accountant for this?
You do not need one to count a drawer, and this sheet is not a tax document. GST treatment on the invoices behind these sales — the CGST and SGST or IGST split, the HSN or SAC codes, the place of supply — is a separate matter, and this worksheet does not record any of it. Confirm your own obligations with your chartered accountant; the only thing this sheet asserts is what was expected, what was counted, and the reason for the difference.
Should UPI and card be counted at all?
They are not counted in a drawer, but they are verified — against the UPI collection report for the day and against the acquirer report, which for card settlements usually arrives the next day. The example verifies them line by line and records where the report came from, because “the UPI total looked about right” is not a check. A card discrepancy you find on the same evening is a mis-tagged payment; the same discrepancy found a week later is a reconciliation argument.
Close the day as a comparison, not a reconstruction.
If expected figures only exist once somebody has remembered what the day looked like, the variance column is measuring memory rather than money.