How to combine months of bank statements into one clean spreadsheet

One account, twelve monthly statements, three formats — and somebody (a lender, an accountant, a tax form, or just you) needs it all as one clean spreadsheet. Concatenating files sounds trivial; doing it without double-counting or format chaos is where the hour goes. Here's the workflow, with the traps flagged.

Why this job keeps coming up

• Tax prep — the year's income and expenses in one sortable place.
• Loan and mortgage applications — affordability reviews across months.
• Catch-up bookkeeping — reconstructing records from statements as source-of-truth.
• Spending analysis — cadence questions (subscriptions, fee patterns) only answer across months.

Step 1 — one format for everything

Gather every statement for the period, then normalise: native CSV exports where the bank offers them, converted PDFs where it doesn't. The converter takes PDF, CSV, Excel, OFX/QFX and text and emits the same column layout for all of them — which is the property that makes step 4 painless. Reconcile each converted file against its statement's own opening/closing balances before moving on ( why and how).

Step 2 — normalise columns and dates

If you're combining by hand, three conventions must match across files before you paste them together: date format (one locale, ideally ISO YYYY-MM-DD, which sorts correctly as text), amount convention (one signed column, negative = money out — convert any debit/credit column pairs), and header row (exactly one, at the top). Mismatched conventions don't error — they silently corrupt, which is worse.

Step 3 — dedupe the overlaps

The step everyone skips, and the one that ruins totals. Monthly statements overlap at the edges; re-downloaded files duplicate whole months; and a "combined" sheet that double-counts inflates income and expenses plausibly — the error survives eyeballing. Dedupe on (date, amount, descriptor) after sorting by date. Watch the one legitimate exception: genuinely repeated same-day identical charges (two coffees) — check candidates rather than bulk-deleting.

Step 4 — identify and categorise

A merged file of raw descriptors still isn't usable — SQ *BLUE BTTL can't be summed into a category. Merchant identification is what turns the sheet into analysis: name, category and confidence per row, after which pivot tables (or an accountant) do the rest. The manual method works for dozens of rows; for a year of statements it's the reason the automated path exists.

Manual vs automated, honestly

Manual (convert one by one, paste, dedupe by formula, identify by search): free, and fine for two or three clean CSV months. Budget an afternoon for a year with PDFs, plus real error risk at steps 2–3. Automated: upload the statements — a zip of the whole year in one pass on paid plans — and the pipeline runs conversion, cross-statement dedupe, identification and subscription/fee flagging, ending in one export: Excel/ CSV for spreadsheets, QIF/OFX for accounting tools, with a combined multi-statement overview on top. If the destination demands a specific shape (a firm's date format, custom columns), the AI transform reshapes it with a verifier proving amounts and dates survived untouched. The bookkeeping guide continues from here into categorisation and reconciliation.