Short answer: no, and you usually don't want it to. ExpenseBot tracks which Gmail messages it has already processed using a Gmail label (ExpenseBot), not by checking whether the file still exists in your Drive. So if you delete the PDFs out of Drive and re-fire a scan, the scanner will see those emails as "already done" and skip them — you'd be left with nothing.
How dedup actually works
When ExpenseBot processes a receipt email, two things happen:
- The structured data lands in your Google Sheet
- The Gmail message gets the
ExpenseBotlabel
On every subsequent scan, ExpenseBot's catch-up query is essentially "find receipts that don't already have the ExpenseBot label." Deleting the Drive PDF doesn't unlabel the email — so the scanner correctly treats that email as already-processed and moves on. The dedup is on the email side, not the Drive side.
What if I want to force re-extraction?
You almost certainly don't, and here's why: re-extracting re-fires our AI on every PDF, which means receipts that are currently parsed correctly could come back parsed differently (or worse). It also requires unlabeling each Gmail message first — there's no "re-process everything" button by design.
The two situations that usually drive "I want to re-scan" requests both have safer answers:
| What you want | What to do instead |
|---|---|
| Apply a new filename or metadata format to older receipts | Ask support about the one-shot backfill. It renames and restamps your existing Drive files using the receipt data already in your sheet — no re-scan, no risk of new extraction errors. |
| You think a receipt was missed | Forward the single email to receipts@expensebot.ai, or manually label it ExpenseBot so the next scan picks it up. Don't nuke everything. |
| You want to start completely fresh | Use the Reset Account option in Settings — that's the supported way to wipe state and start over. Manually deleting Drive files won't trigger it. |
"But I already deleted them — what now?"
If you've already deleted PDFs from Drive expecting a re-scan to bring them back, contact support before doing anything else. We can run a backfill that re-uploads the PDFs to Drive using the receipt data still in your sheet — much safer than unlabeling thousands of Gmail messages and re-firing extraction.
TL;DR
- Delete in Drive ≠ re-extract on next scan
- Dedup is keyed off the Gmail
ExpenseBotlabel, not Drive file presence - For "I want the new format applied to old receipts," ask about the backfill — don't delete-and-re-scan
- For "I want to start over," use Reset Account in Settings