- Home
- Free downloads
- 50 AI Prompts for Bookkeepers
Free download: Prompt pack (PDF)
50 AI Prompts for Bookkeepers
Fifty copy and paste prompts for ChatGPT, Claude, Gemini, Copilot or Grok, grouped by the bookkeeping task they help with.
Each prompt works in any AI chat tool. Replace the text in [square brackets] with your own details, paste it in, and review the answer before anything goes into the books.
Remove client names, account numbers and other personal details before you paste data into an AI tool, and follow your firm's and your clients' data policies. AI output is a draft: you are still the reviewer.
These prompts work best with clean, consistent spreadsheets. The Small Business Bookkeeping Spreadsheet Bundle gives you ready-made sheets to paste into and out of. Small Business Bookkeeping Spreadsheet Bundle: Pricing to Payroll
Our sister store (same team). ReconciledBooksCo is run by the team behind this site, so we earn money if you buy a template there. Affiliate disclosure
What is inside
- Categorize and clean up transactions (8 prompts)
- Bank and credit card reconciliation (8 prompts)
- Month-end close (8 prompts)
- Accounts payable and receivable (8 prompts)
- Reporting and client communication (8 prompts)
- Spreadsheets and formulas (5 prompts)
- Nonprofit and church treasurers (5 prompts)
50 prompts in total. Here are the first 3 so you can see the style.
Categorize and clean up transactions
1. Categorize a batch of transactions
You are helping me categorize bookkeeping transactions. Here is my chart of accounts: [paste account names]. Here are the transactions as date, description, amount: [paste rows]. For each row, suggest one account from my list, rate your confidence high, medium or low, and give a reason in five words or fewer. Do not invent accounts. Put any row you are unsure about in a separate list for me to review.
2. Write bank rules from repeated transactions
Here are three months of bank transactions: [paste rows]. Find descriptions that repeat at least three times with the same likely category. For each, suggest a bank rule: the text to match, whether it should be 'contains' or 'is exactly', the account, and any amount condition. Flag descriptions that look alike but probably belong in different accounts.
3. Clean up vendor names
Here is a list of vendor names from my books: [paste names]. Group names that refer to the same vendor, such as different spellings, store numbers or payment processor prefixes. For each group, suggest one clean vendor name, and show a table of original name and suggested name. Only merge names when you are confident they are the same business.