Bank Statement Analyser
Drop in a PDF statement and every transaction is sorted into a category, totalled by merchant and by day, with repeated payments picked out. It is the same conversion you would run anyway, with the analysis attached; nothing is uploaded and nothing is linked to your bank.
Included in the one-time statement unlock · no subscription
How the analysis works
Convert the PDF as normal; every transaction is read on your own device
Each row is matched against a keyword list and sorted into one of thirteen categories
Totals, merchants, daily spend and repeated payments are worked out from the sorted rows
Export the lot as Excel or JSON, with the categories and figures included
What comes out
- Thirteen categories: groceries, dining, transport, subscriptions, bills, shopping, entertainment, health, travel, income, transfers, cash and everything else
- Spend per category, with each one’s share of the total, so the shape of a month is visible at a glance
- Top merchants by total spend, grouped past the reference numbers banks append to each line
- Repeated payments: anything charged three or more times at a consistent amount, which is how a forgotten subscription looks
- Daily spending, money in against money out, average daily spend and the single biggest expense
- Everything computed locally, so an accountant can run a client’s statement without it leaving the machine
Why not connect an account instead?
Budgeting apps read your accounts through open banking, which means granting a third party ongoing access to live balances and transactions. That is a reasonable trade for continuous tracking, and a poor one if the question is simply where a particular month went, or if the statement belongs to a client rather than to you.
Analysing the PDF answers the same question without a connection, a login or a running permission. It also works on accounts that are closed, on statements from banks no aggregator supports, and on documents someone emailed you.
Statement analysis questions
How are transactions categorised?
By matching the description against a keyword list covering thirteen categories. It is deliberately simple and predictable: you can see why a row landed where it did, and nothing about your spending is sent to a model or a server to decide.
What counts as a recurring payment?
A merchant billing three or more times at amounts within ten percent of each other. That catches subscriptions and standing orders while ignoring a shop you visit often for varying amounts.
Is my statement uploaded to be analysed?
No. Categorisation and the totals run in your browser on data extracted there. Only statements you explicitly unlock and save to an account are stored, and only as extracted transaction data, never the PDF.
Can I get the analysis out?
Yes. The Excel export carries the categories alongside the transactions, and the JSON export includes the full breakdown, merchants and recurring payments for anything you want to do downstream.
Does it cost anything?
Converting and exporting to Excel or CSV is free for ten pages a day. The analysis is part of the one-time statement unlock, around $2.99, with no subscription.