A shoebox and a spreadsheet: only one of them answers a question

Picture the shoebox on the shelf above the desk. Into it goes every receipt, statement, letter, and folded scrap of paper with a number on it. It is not organised. It is not labelled. It is, in the most literal sense, a box.

There's also a spreadsheet. Four columns: date, description, amount, category. In a good month it gets updated weekly. In a bad month it gets updated in a two-hour panic on the thirty-first.

Both of these get called "my system." Only one of them is doing any work, and it's worth being honest about which.

The test that settles it

The question gets forced by something routine — nothing dramatic, a mismatch between two numbers that needs explaining, requiring eight months of a particular category of spending to be reconstructed.

The instinct is to go to the shoebox first, because the shoebox is where the evidence is. Two hours later you are sitting on the floor surrounded by paper, having established that there are roughly a hundred and forty relevant receipts, that a good number of them have faded to blank thermal paper, and that there is no way of knowing whether the set is complete.

Open the spreadsheet, filter on the category, and the answer arrives in about ninety seconds.

The box holds everything and tells you nothing. The spreadsheet holds a fraction and answers the question immediately.

Why the box survives anyway

The shoebox isn't stupid. It comes from a real and reasonable instinct: if someone ever asks you to prove something, you want to be able to. Paper is proof. A spreadsheet row is a claim.

That's true. But it conflates two completely different jobs that almost nobody separates:

• Evidence — being able to substantiate a specific item if challenged.
• Awareness — knowing, roughly, what's going on, in time for it to matter.

The box is a mediocre evidence system and a completely useless awareness system. It stores proof in a form that can't be searched, on a medium that degrades, with no index. The spreadsheet is an excellent awareness system and a weak evidence system — but awareness is the thing you need forty times a year, and evidence is the thing you need maybe twice in a decade.

That is optimising, without noticing, for the rare case over the constant one.

What to change instead

None of this is an argument for throwing the box out. That would be a different essay and possibly a foolish one. It is an argument for changing what each thing is for.

Make the spreadsheet the primary record

Everything of consequence gets a row, on the day it happens, in about twenty seconds. Not a budgeting app, not an elaborate system with rules and automations — elaborate systems tend to get abandoned, because they demand a version of you that doesn't exist on a Tuesday.

Four columns. That's it. Boring enough to actually get done.

Photograph the paper instead of storing it

The thing that breaks the box's monopoly on evidence is that the phone in your pocket is a scanner. A receipt that matters gets photographed the moment it arrives, filed in a folder named for the year, with the amount typed into the filename. That's it. Searchable, backed up, doesn't fade.

Receipts that don't matter can go in the box, where they'll sit undisturbed and never be looked for — which is fine, because that is already what happens to them.

One evening a quarter

Four times a year, sit down for an hour and reconcile: spreadsheet against bank, folder against spreadsheet. It's genuinely dull. It is also what catches duplicate charges, forgotten subscriptions, and categorisation errors that would be genuinely annoying to unpick at year-end.

An hour a quarter is four hours a year. The floor-covered-in-paper afternoon costs two hours by itself and produces nothing.

The general lesson

Most people who describe themselves as "bad with paperwork" are actually doing something quite specific: they're maintaining a system that produces a feeling of diligence without producing usable information.

The shoebox feels responsible. Every receipt that goes in is a small act of conscientiousness. The feeling is real. The output is a box.

The test worth applying to any system you're unsure about: can it answer a question? Not can it store things — can you ask it something specific, and get an answer, in under five minutes, on an ordinary weekday?

If not, it isn't a record-keeping system. It's a place where records go.

There is nothing wrong with keeping the box. It represents years of trying, and that counts for something. The change worth making is to stop pretending it is doing the work.