How to find subscription traps in your bank statement
Subscriptions are designed to be forgettable: a free trial that quietly becomes a paid plan, a price that creeps up a few pounds, a charge that lands once a year and slips past unnoticed. In the UK the problem is now on the statute book — new rules announced for January 2027 will make cancellation easier, require clearer information when you sign up, and add a 14-day cooling-off period after a trial or certain longer-term contracts renew. Until then, the statement is the only place the traps show up, and reading it line by line is the slow way to find them.
This tool does the reading for you. When you paste transaction lines here, your browser:
- Reads each line — date, merchant and amount — and groups the charges by service, so "NETFLIX.COM" and "APPLE.COM/BILL NETFLIX" land in one group instead of looking like two different bills.
- Works out the billing cadence from the gaps between charges — weekly, monthly, quarterly or yearly — and totals the monthly and annual burn at the current amount.
- Checks five trap patterns — duplicate charges, trial conversions, price jumps, the same service billed through two routes, and annual renewals — and shows the exact charges and dates behind every flag, with a trap score per service.
What makes this detector different
The pages that top the results for this search are mostly guides that tell you to audit your statement by hand, or apps that need a bank connection — open-banking tools and subscription managers that link your accounts to find recurring charges, or file-based checkers that need a statement exported as CSV, XLSX or OFX first. None of them is a dedicated free fully client-side tool that takes lines copied straight from a banking app — no export, no upload, no login — and flags the five trap patterns with the evidence shown. This page is the only one we found that combines paste-first input, grouping by service across billing routes, duplicate and trial-conversion detection, a per-service trap score with a plain-text report you can copy or download, and a live network-request counter so you can verify the privacy claim yourself. It runs entirely in your browser — there is no server to upload to.
When is this the right tool?
Whenever a statement feels heavier than it should: a month where the same service appears twice, a charge that looks bigger than last month, a free trial you signed up for months ago, or a yearly charge you only notice when it lands. Paste three months of transactions for monthly traps, and twelve months or more to catch the annual ones — most banks let you copy or export a full year at once. If you already know your subscriptions and just want to total them and decide what to cancel, the Subscription Audit covers that side — this tool covers the traps you do not know about yet.
Honest limitations
A flag is a prompt, not proof
A flagged charge is a reason to check — not proof you are being overcharged. A duplicate-looking charge can be a legitimate retry after a failed payment, and a price rise can be one you agreed to. Read the evidence on the card before acting.
Usage is invisible
The tool sees charges, not use. It cannot tell whether you still watch a service or open an app, so a steady monthly charge with no trap pattern is listed for review rather than flagged. Only you know whether it is worth keeping.
One-off charges stay invisible
A single charge with no repeats cannot be judged — there is no pattern to compare it against. The tool needs at least two charges for a service, and three or more for the trial-conversion check.
Dates are read day-first
Lines written 14/11/2026 are read as 14 November (UK style), not 11 April. If your bank writes month-first dates, the cadence and duplicate checks can misread them — the charge amounts and merchant grouping are unaffected.
Amounts without a symbol use your pick
A line like "NETFLIX 15.99" has no currency in it, so it is read in the currency you pick above the text box. Mixed-currency statements are handled per line, but totals are shown separately per currency and never converted.
Service names come from a fixed list
About forty well-known services are recognised by name, so billing descriptors for the same service group together. A service outside the list still works — it is grouped under its own merchant name — but two different descriptors for it will not be recognised as the same service.