Product guides

A quick rule born from a single transaction

When one transaction shows a pattern worth automating, create a rule straight from it instead of starting from a blank editor.

Editorial illustration for A quick rule born from a single transaction

Some of the best rules are noticed rather than planned. You open a transaction, correct its category, and realise that this will happen again next week. Rather than going to the rule editor and rebuilding the condition from scratch, you can create a quick rule directly from that transaction. The rule starts pre-filled with what the transaction already tells you.

Why start from a transaction

A blank rule editor asks you to decide what to match before you have looked at any examples. Starting from a transaction reverses that. The merchant, description and amount are in front of you, so the condition is based on real data. You edit rather than invent, which reduces mistakes and saves time.

It also keeps rules honest. Rules created from real corrections address real friction. Rules invented in the abstract often solve problems you do not have.

What to adjust before saving

A pre-filled rule needs a short review before you accept it. Three things deserve a second look.

  • **The condition.** A transaction description can contain parts that change every time, such as dates and references. Keep the stable part, usually the merchant or processor name, and drop the rest.
  • **The category.** Confirm the rule files under the category you intend, not just the one currently on the transaction.
  • **The scope.** Decide whether the rule should also apply to existing transactions.

The preview helps with all three. If it shows only the transaction you started from, the condition may be too specific. If it shows many unrelated entries, it is too loose.

Making the condition robust

Imagine a payment described with a business name followed by a reference number and a date. A rule built on the whole string will match exactly once. Remove the changing parts until what remains is something that every payment to that business shares. Then the rule works for the next payment, and the one after.

Be careful not to cut too far. A condition that reduces to a very short common word could catch unrelated payees. After trimming, read the preview again, scanning for names that do not belong.

When a rule is overkill

Not every correction deserves a rule. A genuine one-off, like a gift bought from a shop you rarely use, does not. Ask whether you expect the same correction at least three more times. If so, write the rule. If not, make the single change and move on.

Merchant defaults are a lighter alternative. If the pattern is simply this merchant belongs in that category, setting the default achieves the same result without the machinery of a rule. Reach for a rule when you need extra conditions or other actions.

SituationBest tool
Merchant always belongs to one categoryMerchant default
Merchant belongs to one of two categories depending on amountRule with an amount condition
One-off exceptionEdit just that transaction
Pattern across several merchantsRule with match any

Naming quick rules

Quick rules tend to be made in a hurry and named carelessly. Take five seconds to give each a descriptive name, such as Pharmacy to Medical. A list full of rules called Rule 1 and Rule 2 is impossible to maintain, and the day you need to change one you will not know which it is.

A habit that compounds

Each time a correction teaches a rule, the review queue gets smaller. After a few months of turning repeated corrections into rules, most new transactions arrive in the right place. The effort is spread thinly across ordinary use, which is far easier than a single marathon setup session.

Remember that only some roles may write rules in a shared workspace; if the option is missing, ask an administrator.

SPENDIXAI JOURNAL

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

Explore your reports