Import years of history across many accounts
On this page
Say you're moving years of history for three accounts that have no bank link: a checking and a savings account at your credit union, and a Holiday Fund you already track in Pennyweight. You have one statement file for each. Ask your AI to plan the import first, then bring the accounts in one at a time. If you haven't read how importing works, start with moving to Pennyweight.
How to do it
- Hand your AI the files and say which account each one belongs to. Your AI reads each file and turns its lines into rows, whatever the format. See import OFX, QFX or QBO files and import QIF files for format details.
- Ask for the plan before anything is imported. Your AI sends Pennyweight a summary of each account in the files: its name, how many rows it has and its date range, not the transactions themselves. The plan says, for each account, whether it matches an account you already have or needs to be created, and splits its rows into batches of up to 250. Planning saves nothing, and one plan covers up to 100 accounts.
- Create the missing accounts. For an account the plan can't match, your AI adds a manual account, with the balance it had at the start of your oldest statement. The plan tells your AI to show you this step first. See add a manual account.
- Give each existing account its opening balance. An account you already track with no opening balance, such as the Holiday Fund here, needs one before its first import saves. See set or correct an opening balance.
- Import batch by batch. For each batch, your AI shows a preview and asks for your OK before saving. See reading an import preview.
- Check the result. After the last batch, compare each account's balance with your statement. Then work through after you import.
Before importing anything, show me the plan for these three statement files: statement.ofx is my Cedar Credit Union Checking, statement.qif is my Cedar Credit Union Savings, and iso.csv is my Holiday Fund. Which accounts need to be created, and how many batches will it take?
Here are statement files for three different accounts: statement.ofx is my Cedar Credit Union Checking, statement.qif is my Cedar Credit Union Savings, and iso.csv is my Holiday Fund. The two Cedar accounts aren't in Pennyweight yet. Balances at the start of the month were $1,000.00 for checking, $2,500.00 for savings and $300.00 for the Holiday Fund. Plan the import, add the two new accounts, show me a preview for each file, then ask me once before importing.
What you'll see
For the first prompt, your AI lists the three accounts with their row counts. It says the two Cedar accounts need to be created, the Holiday Fund already exists, and each account fits in one batch. Nothing is created or imported.
For the second, your AI adds the two Cedar accounts with their opening balances and shows a preview for each file. After you say yes, it saves all three batches and reports how many rows went into each account.
Good to know
- Already imported? If an account already has imported history, the plan warns you, with the dates and row count already there. It suggests skipping that account or importing only dates outside that period, so the same rows aren't counted twice. Your AI can also list your past imports.
- Closing balances. If you give your AI each account's closing balance, the plan includes it as the number to check the account against after the last batch.
- Linked bank accounts. For a linked bank account, the plan targets that same account for older history. Import only history older than what the bank link already brought in.
- Batch size. Your AI sends at most 5,000 rows per import and states how many rows each batch should have. Pennyweight refuses a batch whose count doesn't match, so a dropped row can't slip through.
- AI apps that can't save files. Some AI apps can't keep prepared batches between steps. Your AI can hold them in Pennyweight while it works: encrypted, in parts of up to 300 KB, set to expire after 2 days. See import a statement from your phone.
- Your rules run on every batch. When a batch is saved, your active rules categorize the rows they match, and a matching rule outranks the category that came with the file. To keep the file's categories, ask your AI to import without applying your rules. See import a CSV file.
- One batch, one undo. Each saved batch is its own import, so you can remove one batch without touching the others. See undo an import.