I tracked my expenses by hand for eleven weeks, then stopped for four months. Nothing dramatic happened. I travelled for a week, came back to a pile of untyped receipts, and the backlog was heavier than the habit had ever been. That is how most tracking dies. Not from weak discipline, but from a system that asks for five minutes every day and hands back one useful number a month.

The version that stuck is smaller and blunter, and it starts by admitting that most lines on a statement do not matter.

What expense tracking actually needs to answer

Three questions, and only three:

  1. What came in this month, and on what date.
  2. What is already committed before any decision gets made: rent cheque, school fee, loan instalment, insurance, subscriptions.
  3. What is left loose, and roughly where it went.

Everything past question three is curiosity, not control. Knowing that AED 41 of the grocery bill was cheese changes nothing. Knowing that AED 2,300 of a AED 4,000 loose budget disappeared into delivery apps changes the next month.

A typical statement of mine runs around forty lines. Maybe eight of them move the answer to question three. The rest are the same supermarket, the same fuel station, the same coffee shop, in slightly different amounts.

Tracking is not data entry. It is reconciliation.

Manual entry vs importing a bank statement: which one survives month two

Manual entry has one honest advantage. Typing an amount makes it register emotionally in a way that a monthly summary never does. Its cost is that it breaks the first busy week, and a tracking system with a two-week hole is worse than none, because the totals now lie in a direction nobody can estimate.

Importing has the opposite profile. It never forgets, and it arrives messy. UAE statement PDFs are hostile to parsing in specific, repeatable ways: rows split across page breaks, merchant names arriving as card reference numbers instead of shop names, Arabic and English inside the same description field, and fee lines labelled in abbreviations rather than words (“VAT ADJ”, “SVC CHG”, “CHRG”).

Those fee lines carry real money. VAT has applied in the UAE at a standard rate of 5% since 1 January 2018 under Federal Decree-Law No. 8 of 2017, so a AED 25 service charge leaves the account as AED 26.25, usually as two separate rows. Anyone summing a fee column by hand tends to catch one row and miss the other.

The split that works: let the bank and card side import itself, and reserve manual entry for the one thing no statement can see, which is cash.

How to design expense categories that still make sense in month six

The rule I use: a category earns its existence only if a different number in it would change a decision. If seeing AED 300 instead of AED 700 in a category produces no action, that category is a label, not information.

At the trading company where I work, suppliers are not grouped by what they sell. They are grouped by payment terms, because the terms are what decides which invoices go out this week. The product type is trivia. The same logic transfers to a household ledger: “Food” as a single line is useless when half of it is a fixed weekly grocery run that will not change, and half is delivery ordered at 9pm, which absolutely can.

Seven to nine categories, sorted into three shelves, is enough for almost every UAE household I have seen:

The third shelf is the one people skip, and it is the one that produces the “where did the month go” feeling every year in the same three months.

The two-bank problem, and the error that inflates every total

Most people here run a salary account at one bank, a credit card at another, and an exchange house for remittances. That is three ledgers, and the most common tracking error lives exactly between them: recording a transfer as an expense.

Money moving between accounts under the same name is a transfer, never spending. Moving AED 5,000 from the salary account to cover the card bill is not a AED 5,000 expense, and counting it as one double-counts every card purchase already recorded that month. Sending AED 5,000 to family abroad is spending, and it belongs in a category of its own rather than buried in “Other”. The remittance fee and the exchange margin belong there too.

Cash needs less ceremony than people expect. One line a week, “cash spent, AED 400”, assigned to the category it mostly went to, is accurate enough to be useful and cheap enough to actually get written down.

That forty-line PDF is the reason the statement importer in Masrofna exists at all. But the method matters more than the tool: the moment tracking stops being a diary and becomes a monthly reconciliation of numbers that already exist, it stops needing willpower. Eleven minutes once a month beats five minutes a day that never reaches month two.