You stopped tracking expenses. Here's how to start again

The app has been on your phone for four months. The last entry is dated April 23rd.

Every time you open it, you’re met with an empty April, a half-finished May and a question with no obvious answer: where would you even start now. Closing it is easier.

This is a common place to be, and there is a way out. It begins with a decision that feels wrong.

Don’t backfill

The instinct makes sense. Sit down on a Saturday, pull three months of statements, enter everything. Then the data is complete and you can carry on with a clear conscience.

In practice it ends the same way every time. Reconstructing three months takes about four hours; somewhere in the second month you lose interest and stop. Now you have two abandoned attempts instead of one.

There’s a second reason, less obvious. Numbers rebuilt from statements and memory are inaccurate — cash is invisible, transfers are ambiguous, and the purpose of half the transactions is forgotten. But they look identical to real ones. From then on you’re making decisions on figures you invented, with no way to tell them apart from measured ones.

Leave the missed months missing. Tracking exists for decisions, and decisions run on recent data.

Why it stopped last time

A second attempt that skips this question ends where the first one did.

Three causes account for most cases, and they’re easy to tell apart.

The entry took too long. If one purchase needed six fields and twenty seconds, friction stopped you, not character. The fix is removing steps, which is the subject of logging an expense fast.

Too many categories. Fifteen categories means a decision on every entry, and decisions cost seconds and patience. Four is enough.

No payoff. You recorded for a month, never opened the totals, and the point evaporated. The fix here isn’t in the recording at all — it’s adding one regular moment where data turns into a conclusion.

An honest answer to “what actually stopped me” is worth more than any new system.

Restarting takes one evening

Fifteen minutes, and no more than that.

Start by getting the old data out of sight. Don’t delete it, just stop looking at it. An unfinished period on screen works as a reminder of failure rather than as information.

Then shrink the system. Fifteen categories become four. Entry through a form becomes voice or a quick field. The goal of a second attempt isn’t tracking better, it’s tracking more cheaply in effort.

Next, pick one moment a week to look at totals. Five minutes on Sunday, three questions. Without this step the records become a ritual with no output again.

Then log today’s purchase. Not the 1st of the month, not Monday. The next thing you buy, however small.

The first two weeks

The bar for this stretch is lower than seems reasonable, deliberately.

Record only what you remember at the moment of paying. Missed one, don’t chase it in the evening. Lost a day, don’t file it as a failure.

Cut nothing. The first fortnight is noise regardless: you don’t yet know what’s typical for you, so any decision from it is arbitrary.

Don’t add categories, however tempting. Splitting one makes sense once it’s grown large enough to be in the way, not in advance.

These two weeks have one purpose: reaching the point where an entry no longer requires a decision. That usually lands somewhere between day ten and day twenty.

The sign that it will hold

It’s a fairly reliable one. You record a purchase without first considering whether to record it.

While every entry needs a micro-decision, the habit hasn’t formed and any busy week will break it. Once the decision disappears, tracking becomes nearly free, and gaps stop being catastrophic — you simply continue with the next purchase.

Until then, keep the bar low. After that, you can gradually add back whatever seemed necessary before: splitting categories, notes, monthly reviews.

What to do about the gap

In a month you’ll have one complete month with a hole in front of it. That’s fine and it gets in the way of nothing.

Comparing months only works between complete ones anyway, so your first honest comparison arrives after two. The gap affects neither today’s remaining figure nor this month’s category breakdown.

One thing is worth holding in mind: annual and quarterly charges may have landed inside the missed period. Write those down separately from memory, without trying to reconstruct exact amounts, and fold them in as irregular costs in your monthly budget.

What not to do on the way back

Don’t start with a new app. Switching tools feels like a clean slate, but it resets every comparison you had: three weeks here and three weeks there produce no comparable month at all. Go back to whatever you were using, unless the tool itself was the reason you left.

Don’t audit the old data. Habit first, conclusions later. Trying to return and optimise simultaneously loads the second attempt with exactly what broke the first one.

Don’t promise yourself a perfect record. A “no gaps” standard turns the first gap into a reason to quit, which is the most common way this fails. Assuming gaps will happen, and that they mean nothing, holds up far better.

On how large the system should be, how many expense categories you need is worth a look before attempt two, since an excess of categories is the usual source of fatigue.

Today

Open the app, set it to four categories instead of fifteen, and log your next purchase.

The previous months will stay empty, and that isn’t a problem awaiting a solution. It’s simply a period when you weren’t tracking, and it ended today.