To check a new expense tracker against your old records, pick one closed month, compare total spend and each category side by side, and explain every gap before you trust the new tool. Small, explainable differences are normal: transfers, refunds, uncategorised rows and timing. A gap you cannot explain, especially a large one, usually means a missing account or a duplicated row.
Which month should I use?
Choose a month that is finished, with every account imported for the whole calendar month and nothing unusual such as a house move. Close the month in your old records too. If your credit card statement runs from the 20th to the 19th, either align the month to the cycle in both systems or accept a known timing difference.
What do I put in the comparison?
- Total spend and total income for the month in each system.
- The total for each category you care about, usually five to eight.
- The count of transactions.
- The ten largest transactions from each.
- A note of anything you changed by hand in either place.
A worked comparison with invented figures
| Category | Old records | New tracker | Gap | Likely reason |
|---|---|---|---|---|
| Groceries | ₹14,200 | ₹14,200 | ₹0 | Matches |
| Eating out | ₹5,100 | ₹5,640 | ₹540 | Tracker included two UPI payments the old list missed |
| Transport | ₹3,800 | ₹3,800 | ₹0 | Matches |
| Shopping | ₹9,300 | ₹8,100 | -₹1,200 | A refund of ₹1,200 netted in the tracker, ignored in the old list |
| Uncategorised | ₹0 | ₹2,350 | ₹2,350 | Rows not yet categorised in the tracker |
| Total | ₹32,400 | ₹34,090 | ₹1,690 | Sum of the gaps |
Check the arithmetic: the old column adds to 14,200 + 5,100 + 3,800 + 9,300 + 0 = 32,400. The new column adds to 14,200 + 5,640 + 3,800 + 8,100 + 2,350 = 34,090. The difference is 34,090 - 32,400 = ₹1,690, and the gaps add up the same way: 0 + 540 - 1,200 + 2,350 = ₹1,690. The figures are invented, but the habit is the point: every gap gets a reason.
Which differences are normal?
- Transfers between your own accounts. One system counts them as spending, the other does not.
- Refunds. The old list may ignore them, while an imported credit reduces a category.
- Uncategorised rows. They sit outside your categories until you assign them.
- Timing. A payment made on the 31st may post on the 2nd.
- Cash. If your old records have cash that you have not yet entered, the tracker will be lower.
- Credit card bill payment. It should not count as spending if you also imported the card's purchases.
Which differences are a warning?
- The same payment appearing twice. SpendixAI flags possible duplicate imports, and its Possible duplicate charges report lists the same amount to the same merchant within two days. Check the import history as well.
- A debit showing as a credit, which means a row was read wrongly. You review rows before saving, so check them.
- A whole account missing, which shows as a big shortfall.
- A mismatch between a category's total and its largest rows.
How do I fix what I find?
Correct the entry in the tracker, not in your memory. If a row was categorised wrongly, change it, and the correction becomes history that similar merchants will follow. If a row is missing, import the missing statement or add it by hand. If a row is duplicated, remove it. Deleted transactions go to a recycle bin for 30 days, which gives you room to undo a hasty clean-up.
When can I trust it?
When your total spend agrees within explainable gaps and the biggest ten transactions match. After that you do not need to repeat the whole check every month. A short spot-check, total spend against the statement totals, is enough.
Useful ideas start with a clearer picture of your own spending.
Explore your reports


