Why automatic spending categories get it wrong (and how to read them anyway)

Automatic categories fail because banks send very little information. The 5 most common errors — internal transfers being the most distorting — and 4 habits for reading them without auditing every cent.

Written by FinBoard Team Published on 2 min read
A pantry shelf of hand-labelled jars; the one marked "Supermercado" (groceries) holds a pair of headphones instead of food, next to jars of lentils and rice.

Your finance app says you spent €240 on "Groceries" last month. Look closer and you'll find headphones, a phone top-up and a group dinner your friends paid you back for. Automatic categories are useful — and less accurate than they look, because they work with very little information.

What the system actually sees

A card payment usually reaches you as a cryptic line like `CARD PURCHASE 4521 MERCHANT*1234 CITY`. The system knows roughly where, how much and when. It doesn't know what was on the receipt.

The 5 most common errors

  1. The merchant isn't the product. Supermarkets sell electronics; online marketplaces sell everything.
  2. Payment intermediaries hide the real shop.
  3. Transfers have no context. A €30 instant transfer could be your share of dinner, a loan repayment or a birthday gift.
  4. Refunds get separated from the original purchase, inflating one month and deflating the next.
  5. Transfers between your own accounts look like spending and income. Move €500 from your current account to savings at another bank, and unless it's recognised as internal, both your spending and income look €500 higher — while your total wealth hasn't changed. This is the biggest distortion when you combine several banks.

Four habits

Review only your 10 largest transactions; spot internal transfers first; watch trends, not exact figures; keep categories few (8-10 beats 40).

FinBoard categorises and tracks transactions across all your accounts in one place, but no automatic categorisation is always right — ours included. Use categories to see patterns, not to audit every cent.

Related posts

← Back to blog