Detecting Duplicate Statements and Transactions
DocuClipper catches duplicate bank statements at two levels - files you already converted, flagged at upload before any pages are used, and overlapping transactions across a project - so re-uploads, reissued statements, overlapping periods, and the same account as both PDF and CSV don't double-count or double-charge.
Last updated
When you process a lot of bank statements, duplicates creep in: the same PDF gets uploaded twice, a bank reissues a corrected statement that overlaps the original, or you import the same account once as a PDF and again as a CSV. Left unchecked, duplicates inflate totals, break reconciliation, and waste pages.
DocuClipper detects duplicates at two levels - when files are uploaded, and across the transactions already in a project - so you catch them before they distort the numbers.
File-level detection at upload
Before DocuClipper spends any pages, it checks whether you have already converted the same file. This works on the bank statement, credit card, brokerage, check, invoice and receipt, and tax form upload pages, whether or not you upload into a project.
As soon as your files finish uploading, any file that is identical to one you converted before is listed in a notice:
- already added - the file is already in the project you're uploading into, with the date the earlier copy was added.
- converted - the file was converted somewhere else: in another project, or not in any project. The notice says when and where, and Open results takes you straight to that earlier conversion, so you can download its data without converting it again.
You then pick one of two options before you click Convert:
- Skip duplicates (selected by default) - only the new files are converted. The page counter shows just the pages that will be converted, and skipped files use no pages.
- Convert anyway - every file is converted, including the ones you converted before, and they use pages again. Inside a single project this also adds a second copy, so project exports count its transactions twice.
The preview on the right opens on a file that will be converted, so with Skip duplicates selected you see one of your new files rather than a copy that is skipped. You can still pick any uploaded file from the Document list.
If every file you uploaded was already converted, there is nothing new to convert: Convert stays disabled until you add a new file or choose Convert anyway. Use Open results to get to the data you already have.
On a team, the check also covers files your colleagues converted, as long as you can open their results: a contract admin sees every member's conversions, and other members see conversions in projects shared with them and unfiled conversions of their team. A file converted by another DocuClipper account never counts.
Files that arrive through automations (email forwarding and cloud folders) are checked against the project they are sent to, without asking: an identical copy of a file already in that project is skipped, uses no pages, and is not added to the project again.
An earlier copy counts only while its conversion exists: converted, still converting, or waiting for pages. A copy from a conversion you deleted, or one that failed, doesn't count, so uploading that file again converts it as a new file.
The same file uploaded twice in one go is converted once. On the bank statement upload page, the What we found in your files notice names the copy, and also points out two different files that cover the same account and statement period: see upload a mix of documents.
File-level detection only catches files that are byte-for-byte the same (it compares the stored file, not the visible content). A reissued statement, a re-scan, or the same account exported in a different format won't be identical - those are caught at the transaction level.
Transaction-level detection across a project
Even when two files aren't identical, they can still cover the same money. DocuClipper looks across all the statements in a project and flags likely duplicate coverage. It catches:
- Overlapping periods - two statements for the same account whose date ranges overlap (for example, a Jan 1 to 31 statement and a Jan 15 to Feb 15 statement that share the second half of January).
- Reissued or corrected statements - a bank sends a replacement statement covering a period you already uploaded.
- The same account uploaded as both a PDF and a CSV - you convert the PDF statement and also import the bank's CSV export for the same account and dates.
Detection groups transactions by date and amount, then confirms each pair with description similarity. The account is used to keep unrelated accounts apart, but it's matched loosely - on the last 4 digits, and with CSV placeholder accounts (like account1) allowed to match a real account - so a masked number or a bare CSV export still lines up with the PDF it came from. When two sources line up this way, DocuClipper treats them as potential duplicates rather than two separate transactions.
Only the statements currently in the project are compared, and the check runs each time you open it. To clear a flagged copy, move it to another project with Move documents between projects or delete it: it drops out of this project's duplicates right away, and is compared with the statements of the project it moved to instead.
Where duplicates are surfaced
Transaction-level duplicates show up in two places:
- A post-upload prompt. After processing, if DocuClipper sees overlapping coverage it prompts you so you can review before the duplicates affect any analysis or export.
- The Coverage tab. Overlapping ranges are listed under Overlapping periods (potential duplicates), alongside the coverage grid for the project. This is the same place you confirm continuous statement coverage, so you can spot both gaps and overlaps in one view. The Coverage CSV export adds a Missing Periods column per account, so gaps and duplicates come out of the same report.
From there you can remove the redundant statement so totals and reconciliation reflect each transaction once.
Why this matters
- Reconciliation. A duplicated statement double-counts credits and debits, so a statement that should reconcile won't.
- Cash-flow and underwriting. Overlapping periods overstate deposits and inflate averages - a borrower's "monthly deposits" jump if January is counted twice.
- Forensic review. Duplicate uploads are sometimes used to pad a document production; surfacing overlaps makes that obvious.
FAQs
DocuClipper says a file is "already in this project." What does that mean?
The file you uploaded is identical to one already in the project. Keep Skip duplicates selected and click Convert to convert only your new files; the copy uses no pages. If you meant to replace the earlier copy, remove it from the project first (see Managing Project Documents) and upload again.
DocuClipper says a file "was already converted." Do I need to convert it again?
No. You (or a colleague) already converted that exact file, in another project or outside any project. Click Open results next to the file to open that conversion and download its data, and keep Skip duplicates selected so the file uses no pages. Choose Convert anyway only if you want a second, separate conversion of the same file.
I sent the same statement to my project's forwarding address twice. Was it converted twice?
No. Email forwarding and cloud-folder automations skip a file that is identical to one already in the project, so it is converted and charged once.
I uploaded the same account as a PDF and a CSV. Will the transactions be counted twice?
DocuClipper detects that overlap at the transaction level (grouping on date and amount, confirming with description similarity, and matching the account on its last 4 digits) and flags it as a potential duplicate in the Coverage tab and in the post-upload prompt. Review the overlap and remove whichever source you don't want to keep.
Where do I see overlapping periods?
Open the Coverage tab for the project. Overlaps are listed under Overlapping periods (potential duplicates), next to your coverage summary. You'll also get a prompt right after upload when an overlap is detected.
How do I remove a duplicate statement?
Remove the redundant statement from the project via the Documents sub-tab. See Managing Project Documents. Reconciliation and any analysis update automatically once it's gone.