Bank statements

How to categorise IDFC FIRST Bank transactions

Categorise IDFC FIRST Bank transactions in four passes: transfers, repeat merchants, people and one-offs, with rules so the next month needs far less work.

Editorial illustration for How to categorise IDFC FIRST Bank transactions

The fastest way to categorise IDFC FIRST Bank transactions is in four passes rather than line by line: first the movements that are not spending, then repeat merchants, then payments to people, and finally the one-offs. Each pass ends with a rule, so by the following month most rows are sorted before you open them.

What does a first-month cleanup look like?

Say your first import brings in 120 lines covering UPI, NEFT, IMPS, ATM, card swipes, standing instructions and charges. The split below is hypothetical, but the shape is typical: a few types of rows account for most of the work.

PassWhat you settleLines handled (hypothetical)
1Transfers between your own accounts and credit card bill payments8
2Ten repeat merchants turned into rules61
3UPI payments to individual people19
4One-offs, decided by hand32

The lines add up: 8 plus 61 plus 19 plus 32 is 120. Notice that ten rules cleared more than half the month, while the fourth pass is the slowest, and it will shrink as the rules multiply.

How do you read the narration?

These sample lines are invented in a common style and are not taken from any IDFC FIRST Bank statement. Your own narration will differ, so use them only to practise the idea.

  • UPI/DR/508812345678/FRESHMART/okaxis becomes Groceries, keyed on FRESHMART.
  • IMPS/P2A/SUNITA K/HELP becomes a person; decide whether it is domestic help or something else.
  • POS 4455XXXX9921 FUEL STOP/PUNE becomes Fuel, keyed on FUEL STOP.
  • SI-OTT PLAN-MONTHLY becomes Subscriptions, a recurring bill.
  • NEFT CR-EMPLOYER-SALARY becomes Income.

Which passes should come first, and why?

Start with the rows that are not spending, because they distort every total. A transfer to your own other account and a payment of your credit card bill look like big expenses and are not. Once they are labelled, your totals are smaller and more believable, and any later pass works on a cleaner picture.

What do you do with payments to people?

This is the hardest group, since a name carries no category. Ask three questions. Is it regular? A monthly amount to the same person is likely rent, help or a fee. Is it round? A round amount often means an instalment or a shared bill. Do you remember it? If not, scroll to messages from that day. Then write a rule for the person only if the purpose is always the same.

What if one shop appears under three spellings?

It will, because card terminals and UPI apps shorten names differently. Say your grocer shows as FRESHMART, FRESH MART BANER and FM GROCERS. Make one category rule for FRESHMART, then add the other two as aliases of the same merchant. All three then land in Groceries, and the next odd spelling is caught by similarity to the first. Do this only for names you have actually seen twice, otherwise you build a long list of rules that never fire.

How does this work in SpendixAI?

Your own rules run first, then your history and similar merchants, and only then, if you have turned it on, an AI model for the rest, working from the normalised merchant name and an amount band rather than your name or account numbers. If AI is off, the leftovers are held for review. Every correction you make becomes history for the next import.

What should you check at the end?

  • No row is left as Uncategorised, or the leftover count is small and known.
  • Transfers and card bill payments are outside your spending view.
  • Charges and interest have their own categories.
  • Two spellings of one shop share a single category, using an alias where needed.
SPENDIXAI JOURNAL

Useful ideas start with a clearer picture of your own spending.

Explore your reports