A catch-all email address sits on a domain whose mail server accepts every message, whether the mailbox exists or not. A verification tool cannot confirm it, so it comes back as "catch-all" or "unknown" instead of valid.
They are common. ZeroBounce, a verification vendor, says catch-all results made up more than 9% of its 2025 verification volume, over 1 billion of about 11 billion addresses, and that 30% or more is normal in enterprise-heavy lists.
You can send cold email to catch-all addresses, but never mixed into your main send. Pull them into their own segment, send a small first batch, and let the bounces decide whether the rest goes out.
My ceiling for a whole campaign is a bounce rate under 0.5%. Below is the math on how fast an untested catch-all segment breaks that, and a rule for when to send, test or drop.
What is a catch-all email address?
A catch-all is a setting on the whole domain. The company tells its mail server: anything addressed to @ourcompany.com, deliver it somewhere, even if the name before the @ matches no real mailbox.
It is an ordinary admin option. Google Workspace documents it as a routing setting that redirects mail for "All inactive and unrecognized accounts" into one designated mailbox (Google Workspace Admin Help, checked September 2026). IT teams switch it on so mail to a misspelled name or a former employee is not lost.
So "catch-all email" really means "an address at a catch-all domain." The person may be real. The address may be a guess. From the outside, the two look identical.
Why can't a verifier tell if a catch-all address is real?
A verifier asks the recipient's server, during the SMTP handshake, whether it would take mail for that exact address. No message is sent. A normal server refuses a mailbox that does not exist. A catch-all server says yes to every address on the domain, so the probe cannot separate one mailbox from another (ZeroBounce).
The tool gets a yes that means nothing. That is why decent verifiers give it its own status, usually "catch-all", "accept-all" or "unknown", instead of calling it valid.
The damage happens in how people read the report. A catch-all server takes everything at the door and sorts it out afterwards, so the report looks clean. Then a dead address either bounces once the server makes up its mind, or the message quietly goes nowhere. Either way you spent a send and learned nothing.
That matters because bounces land on your sending domain, and a bad batch can undo weeks of careful work on the domains and inboxes behind a campaign.
How many B2B emails are catch-all?
Every figure below comes from a vendor, and only one gives both a date and a sample size. Here is what holds up, checked September 2026.
| Figure | What it counts | Source | Date and method | How far to trust it |
|---|---|---|---|---|
| More than 9% of addresses (over 1 billion of about 11 billion) | Individual addresses run through one verifier | ZeroBounce | 2025 volume, article dated July 7, 2026 | Best dated figure available. Vendor's own traffic, no split by list type |
| 30% or more of a list | Addresses in enterprise-heavy lists | Same ZeroBounce article | Stated as typical, no sample given | Directional |
| About 30% of B2B mail servers | Servers set up as catch-all | Dropcontact, a contact enrichment vendor | No date, no method | Order of magnitude only |
Domain rate and address rate are different numbers
The 9% and the 30% do not contradict each other. One counts addresses across everything a verifier processed. The other counts servers, or lists built around large companies.
That difference matters for your own list. If you sell to larger companies, a handful of catch-all domains can each hold dozens of your contacts. Your address share then runs well above the share of catch-all companies.
For a B2B founder, the practical read is this: expect somewhere between a tenth and a third of a verified list to come back catch-all, and more if your buyers work at big companies. Plan for that segment before you plan volume. If the file is months old, verify it again first, because B2B contact data decays while it sits.
How much can catch-all addresses raise your bounce rate?
No verifier can tell you what share of catch-all addresses are dead, because the probe cannot see it. So I run it as a range.
The table shows the bounce rate the catch-all segment alone adds to a whole campaign if every dead address bounces. It assumes the verified part of the list bounces at zero. It will not, so real life is worse.
| Catch-all share of your list | 5% of them dead | 10% dead | 20% dead |
|---|---|---|---|
| 10% | 0.5% | 1.0% | 2.0% |
| 20% | 1.0% | 2.0% | 4.0% |
| 30% | 1.5% | 3.0% | 6.0% |
| 40% | 2.0% | 4.0% | 8.0% |
Put real numbers on one cell. A 2,000-contact list comes back 20% catch-all, so 400 addresses. If 1 in 10 is dead, that is 40 bounces on the first step of the sequence. 40 / 2,000 = 2.0%, from the catch-alls alone.
At my 0.5% ceiling, 2,000 sends allow 10 bounces. The untested segment blows through that 4 times over.
Softer standards do not save it either. A sales software vendor's own 2026 benchmarks call under 3% good and under 1.5% best in class, and say to stop and diagnose above 5% (Amplemarket, published June 23, 2026). The bottom-right corner of the table is past that stop line before anything else has gone wrong.
If you already sent a file like this and the bounces spiked, check your sender reputation before the next send goes out.
The formula that sets your pass mark
Flip the math around. If catch-alls are a share S of your list and your campaign ceiling is C, the catch-all segment can bounce at most C / S. Even that assumes the rest of the list bounces at zero.
- Ceiling 0.5%, catch-alls 10% of the list: the segment can bounce at most 5%.
- Ceiling 0.5%, catch-alls 20%: at most 2.5%.
- Ceiling 0.5%, catch-alls 30%: at most about 1.7%.
That number is the pass mark for your test batch. The bigger the catch-all share, the cleaner that segment has to be.
Should you send to catch-all emails at all?
Usually yes, if they are the right people. Dropping every catch-all domain throws away whole companies, often the larger ones on your list. Dropcontact makes the same case and adds that conversion matters more than bounce rate. I agree with keeping them. I do not agree that bounce rate stops mattering, because it is the number that protects the domain every other campaign sends from.
| Situation | What to do |
|---|---|
| Named decision-makers at target accounts, on a catch-all domain | Keep them. Move them into a separate catch-all segment |
| First send to that segment | Send a small batch on its own. At 100 addresses, every bounce equals 1 percentage point |
| Test batch bounces at or under your C / S limit | Release the next batch, still kept apart from the main send |
| Test batch bounces above the limit | Stop. Suppress the rest of the segment or find those contacts through another source |
| Address built from a guessed pattern on a catch-all domain | Riskiest tier. Send it last, or not at all |
| Cheap or scraped list that shows almost no catch-alls | Check the verification itself before anything is sent |
A 100-address batch is small. The gap between 2 and 3 bounces is partly luck. Treat the first batch as a gate that catches a bad segment. It is too small to be a precise measurement, so if the result lands right on your limit, send a second batch of 100 before you decide.
Role addresses like info@ or sales@ are a separate problem. The mailbox exists, so it usually verifies. The trouble is who reads it: often a shared queue, rarely the person who signs. Filter those on targeting grounds before verification even comes into it.
How I treat catch-alls before a campaign goes out
Catch-alls are where most list checks go wrong. A founder sees a verification report with no red in it, treats "unknown" as good enough and loads the whole file. That is not a pass.
My rule is plain, and it is the same one behind how I check any B2B lead list. Every address runs through NeverBounce or DeBounce before anything is sent. The whole campaign stays under a 0.5% bounce rate. Catch-alls get their own lane: separate, small in volume, watched closely. Only a clean result lets that segment join the main send.
One thing surprises founders. A supplier sends over a B2B file where nearly everything verified and the catch-all column is close to empty. That makes me more suspicious, not less. Clean results like that are rarer than list sellers suggest, so I treat it as a reason to re-check the file. Cheap freelancer lists are the other warning sign I watch for: low opens, high bounces.
My team and I apply that standard to every list we send. Smirnov Consulting Group is a Prague-based B2B outbound lead generation agency that runs cold email and LinkedIn campaigns for founder-led B2B companies and books qualified sales calls.
Questions founders ask about catch-all addresses
Is a catch-all address the same as an invalid one?
No. Invalid means the server rejected that address outright. Catch-all means the server accepted it without checking, so the address may be perfect or may not exist at all. Treat it as untested. That is why it gets its own segment and a small first send, rather than deletion or a full send.
Do people on catch-all domains actually reply?
The catch-all setting belongs to the company's mail server. It says nothing about the person, so a real decision-maker on a catch-all domain reads email like anyone else. The risk sits only in the unconfirmed addresses, and the test batch is how you find out which kind you are holding.
How do I know if my verifier's catch-all count is believable?
Compare it with what vendors report: over 9% of all addresses in ZeroBounce's 2025 volume, and 30% or more in enterprise-heavy lists. A B2B file with close to zero catch-alls deserves a second look. So does any list that sat unchecked for months, since an old result no longer describes the file you are about to send.