Find where your household spending has crept up with Claude

← All use cases / Money

ToolClaude

A dad uploaded two years of credit card and bank statements to Claude and asked it to sort each merchant into spending categories, checking with him on unclear ones. It showed their everyday spending had risen 28% in a year, almost all of it on food, partly a growing toddler and partly a new favourite café nearby.

Prompt to copy
I'm uploading [NUMBER] months of bank and card statements (with account numbers removed). Sort every transaction into sensible categories, ask me about any merchants you're not sure of, and ignore transfers between our own accounts. Then compare this year with last year by category, show me where spending has grown the most, and suggest two or three realistic places to cut back.

Replace the [PLACEHOLDERS] with your own details before you send it.

How to try it

  1. Download your statements as CSV or PDF and remove account numbers first.
  2. Let Claude categorise them, answer its questions about unclear merchants and check a few totals yourself.
  3. Repeat each month with the new statements to spot creeping costs early.

Read the full guide →

More money use cases

Challenge your property tax assessment with ChatGPT’s help

A Maryland homeowner fed his assessment worksheet and the local comparable-sales list into ChatGPT, which picked out where his valuation looked high and drafted the text for a written appeal. The first appeal cut his assessment by $25,000, a modest saving of about $362 a year, and he has since appealed again.

Full guide →
Source: Wallet Hacks

Beat a pricey insurance renewal quote with ChatGPT

When her home insurance renewal went up by £157, a mum from Inverness asked ChatGPT what to do. It suggested adjusting the excess, negotiating with her insurer and what to say on the phone, and she found cover £190 cheaper.

Full guide →
Source: Netmums

Audit a huge hospital bill line by line with Claude

After a relative's death, a family used Claude to go through a $195,000 itemised hospital bill. It flagged charges billed both as a whole procedure and again as components, plus coding problems, and their letters helped bring the bill down to about $33,000.

Full guide →

Reference: Tom Mansfield (LinkedIn)