Cleanup and catch-up work is hard to quote because nobody knows which transactions are wrong until someone has read every one. Reading them is mechanical; deciding what to do with them is not. We do the reading and the preparation, and hand back a short list of what needs a decision, which turns an open-ended job into one you can plan. Our note on why cleanup takes longer than quoted explains where the time usually hides.
What we look for
The same few problems account for most of what needs fixing in neglected transaction data, and each one can be checked for on purpose:
- Payees spelled several ways, which splits one supplier across several names.
- Bank descriptions that hide the payee behind card numbers, branch codes and references.
- Duplicates that share no field exactly, such as the same expense arriving by photo and by email. See the folder that double-counts.
- Transfers between the client’s own accounts mixed in with spending, where coding both ends as a cost counts the money twice.
Every change is listed, never applied silently
You get the transaction file with payees normalised and possible duplicates flagged, not deleted, so the totals before and after still match. Beside it is a list of every change, with the original text next to the new one. A correction made without a record looks exactly like the error it replaced, so we keep the record and you decide.
Catch-up periods
For catch-up and historical work the same applies across many months at once: statements converted, transactions cleaned, missing periods listed. If a keyword rule in the existing books has caught the wrong transactions, the pattern is easy to see once the payees are clean; see a rule that matched inside a word. What to recode, and how, stays with you.