Know why the same purchase appears twice
Duplicates often begin with ordinary behavior. You might photograph a receipt at a counter, import the emailed copy later, and then save a PDF from a vendor portal. A faded printout may be captured a second time because the first photo looked unreadable. None of those actions is careless; they simply create more than one path into the same review queue.
A likely match can also be two real purchases. A regular parking fee, a subscription renewal, or two client meals on the same day may share a merchant and an amount. That is why duplicate review should begin with comparison rather than removal. Similarity is a signal to inspect, not proof that one record is wrong.
Compare the details that distinguish a transaction
Start with merchant, date, total, currency, and any visible transaction reference. Then look at the source documents for time, receipt number, line items, or payment detail that helps separate two entries. The more fields that agree, the stronger the reason to investigate. One matching number by itself is rarely enough to make a final decision.
When the documents are clearly the same purchase, keep the clearer version and note why the other was removed from the working set. If one image shows a full total and the other contains the original page context, you may decide to retain the version that makes later review easier. The choice should serve your workflow, not a rigid rule.
Separate review signals from deletion actions
A useful system can surface possible duplicates so they do not disappear into a long list. That prompt creates a moment for attention, but it should not quietly erase a record. Automatic removal can hide the fact that two legitimate purchases happened to look alike, especially when work involves repeated vendors or similar travel expenses.
Give yourself a small review category for possible duplicates. During a weekly session or before an export, open that category and decide one item at a time. If the answer is uncertain, retain both records and add a note. Keeping an unresolved question is usually safer than creating a clean-looking list with an unexplained gap.
Check duplicate questions before the handoff leaves
The best time for a duplicate check is before a group of records becomes someone else's problem. Filter the ready items, scan for repeated merchant-date-total combinations, and compare only the matches that deserve attention. This is a targeted review, not a demand to inspect every row twice. The goal is to prevent avoidable confusion from traveling with the export.
If another person will review the file, make your decisions legible. A short note such as duplicate image removed or similar purchase retained gives context without a long explanation. It also makes it easier to answer a later question because the reasoning is attached to the period when you had the documents in front of you.
Keep a process that improves with each review
When repeated duplicates come from the same intake path, solve the cause rather than performing more cleanup. Perhaps you can choose one place for emailed receipts, or mark an imported PDF before you also photograph it. A small change in capture habits can reduce the number of false alarms in every later review cycle.
RecordSlip can flag possible duplicates before you export, while leaving the final decision with you. It does not promise perfect duplicate detection and it does not make bookkeeping decisions. It gives your receipt workflow a clear place to compare similar records, preserve useful evidence, and send a more intentional handoff.