Supported files
CSV, XLS and XLSX, up to 20 MB, one file at a time. A P&L, a bank export, a trial balance, a metrics sheet, any spreadsheet with a date column and an amount column, or a debit and a credit column.
Compass reads the file, names what it found, and shows its working as it goes. You can ask it about the file while it reads.
Drop it in
- Open
Numbers, chooseSources, and drop the file in. The first-run screen has the same drop zone. - Compass reads it and shows a card called
How I'll spot duplicateswith the columns that make two rows the same transaction. Use+ Add columnif your file has a better key. - Press
Review rows. Check the mapping underReview fieldsif you want, then choose a destination: add the rows to an existing source or start a new one. - Press the add button, which names the row count.
Import Finishedshows how many rows were added and how many were not counted, with a Review link for anything held back.
Where the rows go
Rows land in Numbers, filed into the source you chose. A source keeps its transactions, and its totals are sums of those transactions, so a re-import changes a total only when the rows change.
Rows that match the duplicate rule are held back and listed in a review panel, where you import or skip each one. A row with the same key and a different amount counts as changed, and you pick Use new value, Keep old or Keep both.
Opening balances, netting entries and year-end closing rows are imported and marked as bookkeeping, so monthly totals show real activity only.
Fixing a column
Before the rows land, open Review fields. Each column has a role such as Date, Account, Amount, Debit, Credit, Description or Ignore, and you change it from the dropdown. A date column plus an amount column, or a debit and a credit pair, is required before you can continue.
Categories that still need an account are listed there too. Pick one, or tick Import and fix later to continue anyway.
After the import, rows that could not be read appear in the review panel with the reason, such as a date that did not parse, and a suggested fix you can apply. If a whole import went wrong, Remove This Import's Rows on the import takes them out, and you can drop the file again.