Classify a bounce export

Parsed entirely in your browser. The file never leaves your device.

How it works

the same 4% bounce rate, two different problems
LIST5.1.1 user unknown · 5.1.2 no such domainVERIFY BEFORE SENDING
YOU5.7.1 blocked · 5.7.26 unauthenticated · 421 rateSTOP AND FIX
THEM4.4.1 timeout · 5.2.2 mailbox fullRETRY LATER
NOISEout of office, auto-replyNOT A BOUNCE
A sending tool reports the rate. It almost never tells you which of these rows you are looking at.
01

Read the export

CSV with any column names, or raw bounce lines. The tool finds the recipient, the sending inbox and the bounce text by header pattern, and falls back to the longest text column.

02

Classify each row

Enhanced status codes like 5.1.1 and 5.7.26, basic SMTP codes, and about 150 phrases receivers actually use are matched in order — auto-replies first so they never count as bounces.

03

Cross-tabulate

Counts by category, by sending inbox and by recipient domain. A single inbox with all the blocks is burned; a single domain with all the blocks is filtering you specifically.

04

Act

Each category carries its action. Export the classified CSV, or just the hard-bounce suppression list to upload back into the sending tool.

The number your tool gives you is not enough

Every sending platform reports a bounce rate, and every operator I work with reacts to it the same way: verify the list harder. Half the time that is the wrong fix. A 4% bounce rate made of 5.1.1 user unknown is a list that was never verified. The same 4% made of 5.7.1 blocked and 421 unusual rate is a domain or inbox that receivers have decided to refuse, and no amount of list cleaning touches it — you will verify a perfect list and bounce on it anyway. This tool exists to split those apart before you spend money on the wrong one.

Where the categories come from

Receivers speak in enhanced status codes — 5.1.1 for a missing mailbox, 5.2.2 for a full one, 5.7.26 for mail that failed authentication, 4.7.0 for rate limiting — and, less reliably, in prose. Google says “the email account that you tried to reach does not exist”; Microsoft says “Recipient address rejected: Access denied” for three different reasons; Proofpoint and Mimecast have their own vocabularies. The classifier matches codes first and phrases second, with auto-replies caught before either so an out-of-office never inflates the count. The bounce code decoder explains any single row in more depth.

Read the inbox table before anything else

When an export includes the sending account, the per-inbox breakdown is the most useful thing on the page. Bounces spread evenly across thirty inboxes are a list problem. Bounces concentrated on three inboxes, and mostly blocks, mean those three are burned and should be paused today, before they take the domain with them. The recipient-domain table answers the mirror question: if one company's mail server rejects everything as policy, that is their filter, and the fix is to stop sending to them rather than to change anything about your setup.

What to do with each pile

Hard bounces come out permanently — the suppression export gives you the list to upload. Soft bounces get a 30-day pause and one retry. Rate limits mean the daily volume is ahead of the inbox's reputation; the warmup calculator gives a ramp that does not trip them. Blocks are the only category that should stop a campaign: pause the affected senders, run the domain checker and blacklist checker on each, and resume at half volume once they are clean.

Frequently asked questions

Which exports does it understand?

Any CSV — Instantly, Smartlead, lemlist, Apollo, Woodpecker, Mailreach, or a raw list of bounce lines. Columns are detected by name, so as long as there is an email and the bounce text somewhere in the row it will parse.

What bounce rate is too high?

Over 2–3% hard bounces on a send tells Google and Microsoft the list was not verified, and repeated sends like that get the domain filtered. Blocks are worse at any percentage: they mean the receiver has already decided about you.

Why are auto-replies in the export at all?

Some tools log every non-delivery response as a bounce, including out-of-office messages. The classifier removes them first and reports percentages against real bounces only.

Is the file uploaded anywhere?

No. Parsing and classification run in your browser; the export never leaves your device.

A row is classified wrong — what then?

The full table shows the raw message next to the category so you can spot it. Rules match codes before phrases, so an unusual receiver message with a standard code will still be right; if you find a pattern I have missed, send it to me.

Last reviewed

Related tools

What to run next

The checks that most often follow this one.

Delivery

More in this category

Read more

Guides that go deeper

Services

When the tools tell you something is wrong

The diagnostics here are free and always will be. When the fix is bigger than a DNS record, this is the work I do.

Get in touch

Start with a call

Bring a domain and the symptom. I will tell you what is actually wrong and whether you need me at all — plenty of people leave that call able to fix it themselves.

Thirty minutes, no pitch

We will run the checks together on your actual domains, and you will leave knowing what is broken, what it takes to fix, and what it should cost. If that is a job you can do in-house, I will say so.

Based inRangpur, Bangladesh — all time zones
RepliesWithin one business day
LicensingWorkspace below list price
Back to top