Importing Transactions

Rather than looking for a specific bank, Cadence Money inspects the header row and matches your columns using flexible heuristics. As long as your file is a standard CSV with one transaction per row, it can almost always read it, even when the column order is unusual.

Moving from another budgeting app is a first-class case. A file whose layout is not recognised on sight still imports fine; you may just need to confirm a column or two on the mapping step, which is the same thing you would do for an unusual bank export.

The wizard is six steps, and it tells you which one you are on at the top. Nothing is written to your account until the last one.

1. Upload. Open the + menu, choose Import Transactions, pick the kind of file you are bringing in - Credit Card, Bank Statement, Budget App or Other CSV - then drop the file in or browse for it.
2. Account. Say which account these transactions belong to. The dropdown offers the accounts you already hold plus Create a new account at the top, and those are the only two answers - there is no free-text name. Picking Create a new account opens the account form on the step itself, and the button becomes Create & Continue. If you have no accounts yet, the step opens on that form already.
3. Mapping. Check which column is which. In most cases Cadence Money has detected everything already.
4. Review. Look through the parsed transactions and check the dates, amounts and merchant names. You can correct a category here, and doing so teaches Cadence Money for next time.
5. Duplicates. Cadence Money compares the file against what it already holds for that account and reports what it found.
6. Confirm. A summary of what is about to land, and the button that writes it.

Only accounts you keep by hand are offered; a bank-synced account is not on the list. Every import belongs to a real account, which is what lets the duplicate check run, what puts the file in Settings → Imports as something you can move or undo later, what lets a rule that names an account match those rows, and what makes the transactions count towards balances and net worth.

When you choose Create a new account, Cadence Money fills the form in from the file before you touch it, and says which part of the file it read: the account named in a column, or the file name. Its guess at the name, the account type and the institution all land in the form, and everything stays editable - a line above it says so.

If the file carries a running balance column, the Balance is filled in too, taken from the newest row and dated to that row rather than to today. That date matters: it is what the next import counts forward from, so a statement that ends three weeks ago is recorded as being true three weeks ago.

The balance is left blank rather than guessed at whenever the file cannot settle it:

• On a credit card or loan. Banks print a balance owed positive in some exports and negative in others, and nothing in the file says which.
• When the account type could not be guessed at all - because the answer you are about to pick might be a card.
• When the file’s currency is unsettled. That is three cases: the file names more than one currency, it names one that is not your default currency, or it names none at all and you hold more than one. In each there is no unit the figure can safely go in. Note that the currency column is mapped on the step after this one, so what counts here is what Cadence Money could already see.
• When the balance is negative - an overdrawn account. The balance field will not accept one, so a prefilled figure would give you a form that quietly refuses to submit.

Type over any of it and it is simply yours: a balance you edited is recorded as today’s figure, not the file’s date.

Cadence Money tries to match each column to one of its standard fields. These are the ones it asks about:

Amount format is the one people miss. Files express direction in three different ways, and the dropdown covers all three: One amount column (expenses negative, income positive), Separate Debit and Credit columns (each row fills one or the other, both positive), and One amount column plus a type column, which then asks you to point at the Transaction Type column that says debit or credit. Picking the split option replaces the single Amount field with a pair called Outflow (expense) and Inflow (income). Choosing the wrong shape is what turns a month of spending into a month of income, so it is worth a look even when the rest of the mapping is right.

That hint changes when Cadence Money is going to flip the column for you. On a Credit Card export - whether you picked that type yourself or Cadence Money recognised one from the file - and on a file whose only money column is an outflow column, the hint under One amount column reads expenses positive, income negative instead. It always describes what the import will actually do.

If your card export already shows charges as negative, the type is the thing to change, not the file: a link under the dropdown takes you back to the upload step to switch it.

There is deliberately no Source dropdown here. Which account a file belongs to is a question about the whole file, so it has its own step; if your export happens to name an account on each row, Cadence Money still reads that column and uses it as the row’s label without asking.

If your file uses non-standard headers, or no headers at all, map each column by hand with the dropdowns. Unmapped columns are ignored, so a file with extra fields you do not care about is fine.

Credit card statements often format charges as positive numbers and payments as negatives, the opposite of a chequing-style export. On the upload step, set the source type to Credit Card to flip the signs, so charges appear as expenses and payments as income.

If your preview looks inverted - income where expenses should be - go back, switch the source type, and preview again.

Duplicates are caught before anything is written. The Duplicates step compares every row in your file against the transactions already stored for the account you chose, and it does so in three passes:

• The original line from the CSV, byte for byte. Every imported transaction keeps the line it came from, so a re-export from the same bank in the same format matches exactly.
• The same date, amount and name, which catches a re-export whose columns moved around.
• A close match - same amount, within three days, a similar description. This is the one for a posting date that shifted or a description your bank rewrote between exports.

The first two are certainties, so Cadence Money resolves them itself and folds them away behind one expandable line that counts them - Show the 14 already imported, with your own number in place of the 14. Only the close matches are put in front of you, each shown beside the stored transaction it matched so you can judge the difference yourself. Everything flagged is set to be skipped; Import anyway on any row overrides that.

Matching consumes a stored transaction rather than just testing whether one exists, so three genuine $4.50 coffees on the same day stay three transactions. The step’s button says Continue, never Import - it moves you to Confirm, and the writing happens there.

A second, separate check runs after the rows land, and it is the one that looks across sources. It compares your new transactions with your synced and manually added ones - same sign, within three days, amount within one percent, similar name - and hides whichever is the duplicate. It also compares imported transactions with imported ones from a different batch on the same account. Anything it hides stays reviewable under Settings → Data & Privacy → Review Duplicates, where you can confirm the match or undo it and bring the hidden transaction back.

The Confirm step can also update the account’s balance for you. It appears as a single tickbox starting Also update, followed by the account’s name and the figure it would write. Ticking it is the whole confirmation - the balance is written when the import is.

The figure comes from one of two places. If your file has a running balance column, Cadence Money uses your bank’s own number from the newest row. If it does not, the figure is calculated from the balance as last set plus every imported transaction since, and the label says how many transactions went into it. Either way it is dated to the newest transaction counted, not to today, because that is the date it is actually true as of.

The offer is withheld rather than guessed at in two cases. On a credit card or loan, a statement figure is left out entirely because banks print a balance owed positive in some exports and negative in others, and nothing in the file says which. And when the transactions are not in the same currency as the account, there is no honest way to add them up, so no figure is offered.

The tickbox also stays away when there is nothing to change - if you created the account back on the Account step and the file’s balance already went in there, the same figure on the same date is not offered to you a second time.

Settings → Imports lists every file you have brought in, with how many transactions it holds, how many of those were hidden as duplicates, and the date range it covers. Two things you can do from that list:

• Move a file to a different account. Pick one from the dropdown on its row. If a batch is spread across more than one account it shows as Mixed, and choosing an account moves all of it. To move only some of its rows, edit them one at a time or use the Account action in Bulk Editing.
• Undo the import. The Undo button removes every transaction the file added, hidden duplicates included. The confirmation names the full number for exactly that reason.

The undo is a genuine reversal rather than a bulk delete. It takes back any goal contributions those transactions made, recalculates the months they touched, and puts back the account balance the import wrote - though only if the account still carries what the import set, so a correction you made since then is not overwritten.

Imported transactions go through the same categorisation as synced ones, and it runs automatically. There is no category-mapping step. Cadence Money works through, in order: any transaction rule you have written, what it has learned from how you categorise that merchant, its own merchant and keyword patterns, and finally the category column from your file if you mapped one. It then links each transaction to a budget based on where it landed.

Rules run ahead of all of that and win, which makes them the way to say “if the name contains X, use category Y” and have it hold. Write one from the + menu’s Transaction Rule entry, or from Settings → Transaction Rules.

Underneath that, Cadence Money learns. One correction is enough: change a merchant’s category once and that choice is applied from then on, including on future imports, and correcting it again only makes it more confident. Every change you make on the Review step feeds that learning.

To start over, Settings → Suggestions has a Clear Learning action. More in Smart Features.

A few patterns trip up imports more than others. If the preview looks wrong, check these first:

• The wrong amount format. If every row imported with the sign flipped, or a whole Credit column was ignored, go back to the mapping step and change Amount format rather than editing the file.
• Cells that are not numbers. Currency symbols, thousands separators and accounting parentheses are all handled, but a cell reading “N/A” or “pending” is reported as a row error and left out rather than imported as $0.
• Day-first dates. Cadence Money switches to day-first when the first number cannot be a month, or when it recognises the export format as day-first. A file where every date is ambiguous - 05/06, 04/03 - has nothing to give that away, so reformatting the column to YYYY-MM-DD is the safe move.
• Blank or summary rows at the top or bottom of the file. A row with no readable date, name or amount is reported as an error and skipped, so deleting them in your spreadsheet keeps the error list meaningful.
• Mixed currencies. Map the Currency column if your export has one. Without it every row takes your default currency, and a file that quietly mixes two is better split in half.
• Statements combining several accounts. Split them apart so each import covers one account and can be pointed at it on the Account step.