Kakeibo vs an app that connects to your bank

What a bank connection actually buys you

Completeness, and it is worth more than people who like ledgers usually admit. Every card purchase, every direct debit, every subscription you forgot you were paying for arrives in the list whether or not you were paying attention that week. The card you use twice a year is in the picture. The annual renewal that goes out on a Tuesday in March is in the picture. A hand kept ledger sees none of those unless you happen to think of them, and thinking of them is exactly what a household with a money problem is not doing.

It also has no week three. Hand entry has a well known shape: the first fortnight is meticulous, something happens in the third week, and the fourth is reconstructed from a statement. A feed does not get tired, does not go on holiday, and does not skip the day you were in a hurry. If the thing standing between you and a record is your own consistency, automation removes the problem instead of asking you to be a better person about it.

What automation quietly removes

The decision. A purchase you enter yourself has to be named and sorted while you are still standing in the shop, and those two small acts turn a transaction into a choice you are aware of making. A purchase that appears on a list on Thursday, already labelled, has never been a choice at all. It is a fact about the past. Kakeibo is not a reporting method and it never was: it is a method for being present at the moment money leaves, and that moment cannot be automated because automating it is what removes it.

There is a second, more practical loss. A category assigned by a rule is a category nobody thought about. The four kakeibo envelopes exist to force an argument with yourself, because the argument is where the useful information is: deciding whether the taxi home was Survival or Optional tells you something about the taxi, and about you, that a merchant code never will. The working list of what goes where, and the cases that genuinely could go either way, are in needs versus wants and the four categories.

The failure modes are different too, and the quieter one belongs to the feed. A hand ledger fails loudly: you stop entering, the pages are blank, and you know. A connected app fails silently, because the data keeps arriving perfectly whether or not anybody looks at it. A year of immaculate automated records that nobody ever read is a very tidy way of learning nothing.

The access question, stated plainly and not scaremongered

Connecting an app to a bank means a third party holds a standing read of those accounts for as long as the connection lives, either directly or through an aggregation service sitting behind the app. That is a factual description of how the category works, not an accusation about any particular provider, and we are not going to make claims about anyone’s security record that we cannot check. What is worth saying is that it is a decision, and it is usually made in about four seconds on a consent screen while you are trying to get to the part of the app you downloaded it for.

Two questions are worth asking before you tap through. Who is on the other end of the connection, the app itself or a service you have never heard of, and what happens to the history if you stop paying or the product closes. Both answers should be on the vendor’s own pages, and a vendor that makes them hard to find has told you something. A hand entered ledger simply does not raise the question, which is a real advantage and also a small one if you were never going to worry about it.

On cost, the category mostly runs on subscriptions, and some products are funded by referring you to financial products instead, which is a different kind of price. We do not print figures for other people’s products here. Free on iPhone and iPad with ads: one savings goal, ten envelopes and twenty five transactions a month. Pro removes the ads and the limits.