A Budget for Everything Fails by Month Two. A Budget for Three Categories Holds.

A Budget for Everything Fails by Month Two. A Budget for Three Categories Holds.

The short version

Most category budgets don't die because the math is hard. They die because they start too wide: a limit on every category turns the first ordinary week into a wall of warnings, and opening the app starts to feel like being told off. The version that survives is smaller and less ambitious — three to five categories, limits set at what you actually spend, one glance a week to see where you stand. We've just shipped Budgets in ExtMoney — a monthly limit per category, with month-to-date spending counted against it — and it's built for exactly that version. Here's the method, with the feature as the tool rather than the point.

Why the budget-for-everything collapses

A budget is a decision tool. It earns its place only where a decision is still possible — where you could genuinely spend more or less depending on what the number tells you.

Put a limit on all twenty of your categories and most of them fail that test immediately. Rent doesn't respond to willpower. Neither does the phone bill or the insurance. Those aren't decisions, they're commitments — in ExtMoney they belong in scheduled plans, which the app already counts against your month, not in budgets. A "budget" for rent is a number that's always exactly met, and its only effect is teaching you to ignore the screen.

The rest fail socially rather than logically. Twenty limits means twenty verdicts every time you look. Nobody consults twenty numbers before buying coffee; what actually happens is that by day ten four categories are over for perfectly good reasons, the warnings stop carrying information, and the whole exercise quietly ends. The all-categories budget doesn't fail on a bad month. It fails on a completely normal one.

Three to five limits, on categories where your behavior actually varies, is not a starter version of budgeting. It is the working version.

Pick the first three from your history, not your ideals

Choosing categories by aspiration — "I should really spend less on subscriptions" — produces budgets on the wrong things. Choose from evidence instead: open your last two or three months of spending by category and look for the ones that are large, variable, and actually decidable. Big enough to matter, different from month to month, and moved by choices you make weekly rather than by contracts you signed once.

For most people that shortlist writes itself: groceries, eating out, and one personal category — clothes, entertainment, a hobby. In ExtMoney the spending-by-category card on the dashboard gives you the ranking in one glance. And the history behind that ranking stays filled in with little effort: categories are suggested automatically from your own past transactions and rules (deterministic code, not an AI reading your ledger), and a transaction itself can be a sentence said out loud or a photographed receipt.

Skip the "guilty" category you barely spend in. A $23-a-month limit on candy makes you feel disciplined and changes nothing. Guilt is not a selection criterion; variance is.

Set the limit at the median, not the wish

Now the number. The classic mistake is writing down the number you'd like to spend. A limit 30% below your actual pattern isn't a plan, it's a New Year's resolution, and it holds for about as long as one.

Take Maya. Her last three grocery months: $438, $391, $412. The median is $412 — so her limit is $410. Not $300 because an article said groceries should be ten percent of income. The first job of a budget is not to shrink the number; it's to make the number visible while it's still moving. Cutting comes later, in $20–30 steps, once a realistic limit has held for a couple of months.

Why the median and not the average? Because one unusual month shouldn't set your limit. Maya's eating-out history is $96, $118, $240 — the last one was a birthday. The average says $151; the median says $118. A $120 limit describes her ordinary month; a $150 limit quietly builds the birthday into every month. This is the same reason ExtMoney's own "Free to spend" forecast uses your median daily spending over the last 90 days rather than the average — one flight inside the window shouldn't distort what an ordinary day costs. (It also needs at least two weeks of history, and says nothing rather than guessing before that.)

Over the limit on the 18th: information, not failure

Some month you'll hit the limit with almost two weeks left. This is the moment most budgets get abandoned — and it's precisely the moment the budget is doing its job. Being over is information arriving early, while you can still act on it.

ExtMoney flags a budget at 80% before marking it over at 100% — on the Budgets screen and on the dashboard card, which deliberately sorts the worst category to the top. When a category does cross its limit, the app notifies you once — one message per category per month, not a daily drumbeat.

At the 80% mark, do the arithmetic; it takes ten seconds. Maya on August 18: $330 spent of $410, so $80 left for 13 days — about $6 a day against her usual $13. Two honest options: slow the category down for the rest of the month, or accept the overshoot consciously and write down why. If she carries on at her usual pace she finishes near $503 — roughly 23% over. Knowing that on the 18th, not on the 1st of next month, is the entire value of the exercise.

What you don't do is delete the budget mid-month out of shame. A limit you remove whenever it triggers is a smoke alarm with the battery taken out.

Transfers and debts are not spending

One design decision worth understanding, because it's where many homemade budgets — and some apps — go wrong: only expenses count against a limit.

Moving $400 to your savings account isn't consuming $400. Paying back a friend isn't a grocery run. If transfers between your own accounts or debt movements landed in a category's total, the number would stop meaning "what this category cost me this month" — and every conclusion drawn from it would wobble. ExtMoney excludes both by construction; there's no setting that can accidentally mix them back in.

The same one-honest-number logic covers shared money. A budget belongs to the whole team: a family sees one limit and one spent figure, not one per person. That's why the limit lives in the team's base currency, and a partner's spending in another currency is converted into it — two people comparing one limit against two different currencies would each get a different verdict, and a shared number can only exist in one currency.

Change the budget, don't break it

A budget is a measuring instrument, not a moral one, and instruments get recalibrated. Two signals that the limit — not you — is what's wrong:

• Over three months in a row, by more than 15–20%. The limit describes a life you don't live. Raise it to the current median, get used to seeing the true number, then cut in small steps if you still want to.
• One category hiding two behaviors. If "Food" is over every month but you can't say whether it's the supermarket or delivery, split it. Two categories with $300 and $120 limits will tell you in one month what a single $420 category never will.

And prune without ceremony: when you retire a category in ExtMoney, its budget goes with it — a limit is an attribute of the category, not a separate thing to clean up.

Practical recommendations

• Start with three. Not by coincidence, three budgets is exactly what the free plan includes; paid plans remove the cap when your system grows.
• Set limits at the median of the last 2–3 months. Realistic first, ambitious later, in small steps.
• Decide at 80%, not at 100%. The warning threshold is your decision point; crossing the limit is just the receipt.
• Read the dashboard card, not your conscience. Worst category on top, month-to-date against limit — most days, that's the whole check-in.
• Recalibrate quarterly. Raise what's chronically over, split what's ambiguous, delete what stopped mattering.

Where we'd push back on ourselves

• Calendar months only, for now. No weekly limits, no custom windows aligned to your payday. If your money month runs from the 25th to the 25th, our budget month doesn't — yet.
• No envelopes, no rollover. An unspent $60 doesn't carry into next month; every month starts clean. If your method is YNAB-style zero-based envelopes, this is deliberately not that — it's a limit, not an allocation.
• Only categorized expenses count. A transaction with no category escapes every budget. Automatic category suggestions and the needs-category counter narrow that gap, but they don't close it to zero.
• The shared currency cuts both ways. A family member who thinks in a different currency reads the budget in the team's currency, not their own. One honest number was the priority; it does mean some mental conversion.
• The feature is new. The decisions above are argued, not yet proven by years of user history.

---

If "three limits, set at what you actually spend, checked in one glance" sounds like a budget you'd still be keeping in November, it's live now: on the web at ExtMoney under Budgets, in the mobile app under Menu → Budgets, and as a card on the dashboard. The free plan includes three budgets — exactly as many as you should start with.

Published 17 August 2026