When this matters
You set up email forwarding (a Gmail rule that auto-forwards Stripe receipts, or you forward invoices manually) and you're wondering: what happens when the rule accidentally forwards a newsletter? Will it create a fake document and burn quota? Will it spam your audit log?
Short version: no. The classifier skips it silently, logs the decision in the Activity log so you can audit, and moves on. This article explains how the classifier makes its call, what each classification does, and what to do when it gets one wrong.
What the classifier checks
Every inbound email is classified with straightforward keyword checks (German and English keyword lists) — deliberately simple and predictable, not a black box:
- Subject and body keywords — "Rechnung", "Invoice", "Receipt", "Beleg", "Quittung", "Statement" and similar terms point to invoice-shaped content. The body is read once for classification and discarded — we don't keep email bodies.
- Attachment names and types — a PDF named like
invoice-123.pdforreceipt-2026-05.pdfis a strong invoice signal. Any PDF attachment is treated as at least a receipt candidate, even if the subject gives no signal. - No matches at all — no invoice-shaped keywords anywhere and no PDF attachment → the email is classified
irrelevant.
The classifier returns exactly one of the four labels. There is no per-email confidence score, and the classifier does not learn per sender — it is a fixed keyword rule set, which is what makes its behaviour easy to predict and to work around (see below).
What happens with each classification
| Class | Action | Counts against quota? |
|---|---|---|
| invoice | Attachments extracted, scanned, document created | Yes — one per attachment |
| receipt | Same as invoice | Yes |
| statement | Same as invoice (bank statements additionally import their transactions) | Yes |
| irrelevant | Logged in Activity log, no document created | No |
The "no document created, no quota use" path is the important guarantee: forwarded newsletters do not silently eat your monthly quota. If you set up an aggressive Gmail rule that forwards a wide net of senders, the false-positive newsletters land on irrelevant and disappear from the pipeline cleanly.
Where to see the classification
Settings → Email integration → Activity log. Every inbound email is listed with:
- Sender, date, subject
- Classification (invoice / receipt / statement / irrelevant)
- Status (processed / skipped / error)
- Links to the documents that were created from it (none for irrelevant)
If you suspect the classifier is missing real invoices, scan the skipped entries for senders that surprise you.
When the classifier gets it wrong
The classifier is tuned for high precision on invoice (don't create fake documents) at the cost of occasionally skipping an edge-case invoice. When that happens, you have two recovery paths.
The invoice got classified as irrelevant
This usually happens with very minimal "see attached" emails — short subject, generic body, no invoice keywords, and an attachment that isn't a PDF (a JPG photo of a receipt, for example — PDFs are always treated as receipt candidates).
Recovery, either of:
- Upload the file directly. Save the attachment and drag it into the web app (or use the mobile app). Fastest and always works.
- Re-forward with a clearer subject. Because classification is keyword-based, re-forwarding the same email with "invoice" (or "Rechnung") in the subject line classifies correctly. The attachment wasn't stored the first time, so duplicate detection won't block it.
The newsletter got classified as invoice
Rarer (the precision tuning prevents most of these) but can happen with marketing emails that carry a PDF attachment (e.g. "Read our latest whitepaper.pdf"). The classifier saw a PDF and treated it as a receipt candidate.
Recovery: open the resulting document and delete it. Documents you move to the trash no longer count toward the current month's document limit, so a stray newsletter PDF doesn't permanently cost you a quota slot. For repeat offenders (the same newsletter keeps tripping the classifier), tighten your forwarding rule so that sender is excluded, or write to [email protected] with a sample subject line.
How quota accounting works here
Quota is consumed when a document is created (after classification + extraction), not when the email arrives. This means:
- Irrelevant emails consume no quota.
- Failed extractions where no document was created (file too large, unsupported format, virus scan) consume no quota.
- Documents you delete stop counting — the monthly counter reflects the documents currently in your account that were created this month.
See understanding your document counter for the full quota accounting.
How reliable is it?
Templated vendor emails (Stripe, Hetzner, AWS and similar) classify essentially perfectly — their subjects and attachment names carry clear invoice signals. The classifier's weak spot is minimal personal emails with non-PDF attachments and no keywords. Because any PDF attachment is treated as a receipt candidate, a real invoice with a PDF attached practically never gets skipped.
If a specific high-volume vendor consistently lands wrong for you, write to [email protected] with a few sample (de-identified) subject lines and we'll look at the keyword coverage.
Edge cases
It classified my invoice as irrelevant. Upload the attachment directly, or re-forward the email with "invoice" in the subject. See the recovery section above.
It classified my newsletter as invoice — there's now a fake document. Open the document and delete it. Trashed documents don't count toward this month's limit.
Can I disable the classifier and force-process everything? No, by design. The classifier protects your quota; without it, every newsletter would consume a slot. If a specific email was skipped, upload its attachment directly. If you want every email processed regardless of content (compliance archiving rather than invoice automation), forwarding may not be the right tool — talk to us at [email protected] about your use case.
I forwarded a calendar invite — what happened? Classified as irrelevant (no invoice keywords, no PDF attachment). No document, no quota use.
I forwarded an email with no attachment. It's classified and logged, but no document is ever created from the email body — only real attachments become documents. If a merchant emails plain-text receipts without attaching a PDF, print the email to PDF and upload it, or configure your forwarding rule to only forward emails with attachments — most providers expose this as a filter condition.
The email doesn't appear in the Activity log at all. It never reached processing — most commonly because sender authentication (SPF and DKIM both) failed, or the address was wrong. See the inbound address explained for sender authentication and email forwarding stopped working for the full diagnostic.
Two-step forwarding (my Gmail forwards to a shared inbox which forwards to TaxItEasy) is showing the wrong sender. Multi-hop forwarding rewrites the From: header on each hop. The Activity log shows the last hop's sender, not the original. For accurate sender attribution, configure the upstream rule to forward directly to TaxItEasy rather than through an intermediate.
Can I see why an email was classified the way it was? The Activity log shows the label and status per email. The keyword checks themselves aren't itemised per email — but they're simple enough to reason about: invoice-shaped words in subject/body, invoice-shaped attachment names, and PDF attachments are the signals.
Related
- Set up email forwarding — the rules that produce the forwarding traffic
- What the AI reads, and how to correct it — the extraction step that runs after classification
- Understanding your document counter — what counts and what doesn't
- The inbound address explained — the address that receives forwarded mail