Paid for 4 tools. Still bounced. So I built my own.

A verifier gives you three answers on a B2B email. Valid, invalid, risky.

Risky means the check failed and you paid for it anyway.

It happens because the domain is a catch all. It accepts every address at SMTP level, including addresses that do not exist, so the verifier cannot separate a real mailbox from a made up one.

I have run a cold email agency for five years. 900+ appointments booked for clients, $2.9M+ in pipeline. Every campaign started with the same step and it was never the copy. It was the list.

Two things kept breaking.

The catch all problem above. The verifier hands back a pile it has no answer for and leaves the decision to you.

Then the gateways. Domains behind Proofpoint, Mimecast, Barracuda, Appriver or Sophos never answer for the mailbox. The SMTP conversation is with the gateway, and the gateway accepts on behalf of the whole domain. Some tools read that as valid. Others give up and mark it risky. Neither one knows.

So I stopped using SMTP as the deciding signal. On catch all domains I resolve through Microsoft Autodiscover and GetCredentialType responses instead, which answer per mailbox rather than per domain. Gateway detection runs first, so those domains get routed to a different check before anything is scored.

That became WizLeads. It ran inside the agency for a year before anyone else touched it.

The only test I have is small and I am not going to dress it up. Live send, 12 addresses my check approved produced 1 bounce. 7 approved by a leading verifier produced 4. Nineteen addresses total. Directionally consistent with five years of bounce reports, nothing more.

What do you do with the risky bucket? Send anyway, drop them, or route them somewhere separate?