Catch-All Emails: What to Do With Them (Using Catch-All Scoring)

Catch-all emails cannot be confirmed, so the right move is to rank them, test the best ones in a small batch, and drop the guesses. Do not mark them valid and do not throw them all away. SimpleVerifier keeps every catch-all row risky with reason "Accepts all addresses" and grades it higher, medium or lower likelihood, with the signals behind the grade. Send higher rows in a test batch, wait for the bounce result, then decide on medium. Lower rows, which are mostly guessed name variants and placeholder addresses, are usually not worth the risk.

How big the problem is

From our own catch-all study of B2B lists, using aggregate check data from 11 to 25 October 2026:

• 30.8% of business domains that gave a definite answer (4,994 of 16,199) accepted a made-up address.
• By mail host: Google Workspace 42.0%, other hosts and security gateways 40.3%, Microsoft 365 17.9%.
• When three or more spellings of the same person's name were checked on a domain that does answer mailbox checks, 1.6% of those guessed addresses existed (70,580 addresses). An address checked in a single spelling existed 37.4% of the time.

That last number is the useful one. On a domain that answers, guessed patterns almost never exist. On a catch-all domain you cannot see that, which is exactly why guesses there are dangerous: they all look deliverable.

What the grades mean

Grade	Typical signals	What to do
Higher likelihood	Same name format as confirmed addresses on the domain; single-word name next to confirmed single-word addresses	Send in a small test batch first
Medium likelihood	Mixed or weaker signals	Hold until the test batch result is in
Lower likelihood	Several spellings of one name in the list; placeholder or no-reply address; disposable domain; random-looking address	Drop, or keep only if you have another source confirming the person

The signals and how each one was tested on 82,853 addresses are on the catch-all scoring page, including the signals we tried and dropped because they did not work. The results file carries catch_all_confidence and catch_all_signals columns so you can sort on them.

A process that works

1. Verify the whole list. Valid rows go to the main send. Invalid rows go on a suppression list.
2. Filter risky rows with reason "Accepts all addresses". Those are your catch-alls.
3. Remove duplicates of one person. If a finder gave you anna.lee@, alee@ and a.lee@ on the same catch-all domain, at most one is real. Keep the one that matches the domain's confirmed format, if any.
4. Test higher-likelihood rows. Send 100 to 300 from a warmed inbox and compare bounces with the rest of the campaign.
5. Decide on medium rows from that result. If the test bounced well under your platform's limit, send medium rows the same way. If not, stop.
6. Leave lower rows out unless something other than the email confirms the person.

Run 100 free checks on a sample of your list to see how many rows are catch-all and how they grade.

What not to do

• Do not trust a tool that returns catch-all addresses as valid. The server accepted a random address too, so that valid is a guess.
• Do not judge by domain alone. A catch-all domain still has real people at it; the grade looks at the address and the list, not just the domain.
• Do not run a catch-all batch from your best inbox first. Use a warmed secondary inbox so a bad batch does not hurt your main sender.

For a deeper decision framework, read should you email catch-all addresses, and for the background, what a catch-all domain is. To check single domains, use the free catch-all checker.