Record every expense in one table with a Payment mode column (UPI, Card, Cash, Wallet), and treat an ATM withdrawal as moving money into your pocket rather than as spending. You count the spending later, when the cash is actually used. A cash-in-hand line then tells you whether your cash entries are complete.
What does the layout look like?
| Column | Heading | Example |
|---|---|---|
| A | Date | 12-Apr-2026 |
| B | Description | Vegetables |
| C | Category | Groceries |
| D | Amount | 180 |
| E | Payment mode | Cash |
| F | Type | Expense |
Add data validation dropdowns for Payment mode (UPI, Card, Cash, Wallet) and Type (Expense, Income, Withdrawal). The Type column is what stops an ATM withdrawal from being counted as spending.
How do ATM withdrawals get counted twice?
Suppose you withdraw ₹5,000 from an ATM. Your bank statement shows a debit of ₹5,000. Later you pay ₹180 for vegetables in cash and ₹300 for an auto ride, and you record both. If you also record the ₹5,000 withdrawal as an expense, the same money appears twice: once as a lump sum and again as the small purchases it paid for. Mark the withdrawal as Type Withdrawal, and leave it out of your spending total.
Monthly spending then uses only expenses: =SUMIFS($D$2:$D$800,$F$2:$F$800,"Expense"). Spending by mode is the same with a second condition: =SUMIFS($D$2:$D$800,$F$2:$F$800,"Expense",$E$2:$E$800,"UPI").
How do you check your cash entries are complete?
Add a small block with four lines. Cash at the start of the period (you count it), plus withdrawals, minus cash expenses recorded, gives the cash you should be holding. Count what is actually in your wallet and the drawer.
- Opening cash, in H1: ₹2,000
- Withdrawals, in H2: =SUMIFS($D$2:$D$800,$F$2:$F$800,"Withdrawal")
- Cash spent, in H3: =SUMIFS($D$2:$D$800,$F$2:$F$800,"Expense",$E$2:$E$800,"Cash")
- Expected cash, in H4: =H1+H2-H3
- Counted cash, in H5: type what you actually count, and put =H4-H5 in H6
With ₹2,000 opening, ₹5,000 withdrawn and ₹4,300 of recorded cash spending, expected cash is 2,000 + 5,000 - 4,300 = ₹2,700. If you count ₹2,550, then ₹150 went unrecorded. That is a good outcome. You now know the size of the gap and can recall what it was, rather than losing it silently.
What about UPI payments that also appear on the statement?
UPI payments from a bank-linked account appear on your statement and in your records. Decide where UPI entries come from and stick to it: either from the statement or typed in as you pay, not both. If you type them and later paste the statement, you will double-count. A wallet is different: topping up a wallet is a transfer from the bank, and what you spend from the wallet is the expense. Treat the top-up like the ATM withdrawal.
If a count comes out higher than expected, you may have received cash too, such as change from a friend or money returned. Add those as Type Income with mode Cash, so the cash check includes money coming in as well as going out.
Is there an easier way to keep both together?
If the mixing is the burden, divide the work: let statements supply the UPI and card rows, and enter only cash by hand. SpendixAI is built that way. You upload statements as PDF, CSV or Excel after choosing the account, so each transaction keeps its account, and you can add cash expenses by hand, one at a time or several at once. It does not read SMS or connect to your bank, so UPI payments reach it through the statement you upload.
A weekly cash and UPI check
- Enter the week's cash spends from your notes.
- Compare expected cash with counted cash.
- Note any gap and what it might have been.
- Confirm UPI rows come from one source only.
Useful ideas start with a clearer picture of your own spending.
Explore your reports


