Enterprise domains look catch-all to email verifiers for three main reasons: a security gateway in front of the mailboxes accepts every recipient and decides later, the domain is deliberately routed to a catch-all mailbox, or (for Microsoft 365) the domain is set to internal relay during a migration or hybrid setup. Plain Microsoft 365 and Google Workspace usually do the opposite and reject unknown recipients at the SMTP stage. When checks return unknown rather than catch-all, the verifier's connection was usually throttled or refused, which says nothing about the address.
This explainer covers how each platform responds to RCPT TO, what gateways change, the signals that still help, and a send policy for these domains. It builds on the pillar, email verification results explained.
How Microsoft 365 responds to RCPT
Microsoft 365 has a feature called Directory-Based Edge Blocking (DBEB). In Microsoft's words, it "lets you reject messages for invalid recipients at the service network perimeter". If the address does not exist, "the service blocks the message before filtering even occurs", with this response:
550 5.4.1 Recipient address rejected: Access denied
For a verifier this is good news: a clear invalid for unknown addresses, and a 250 for real ones. According to Microsoft's DBEB documentation, if all recipients for a domain are in Exchange Online, "DBEB is already in effect".
The exceptions explain most of the "catch-all" results on Microsoft 365 domains:
| Setup | What the verifier sees | Why |
|---|---|---|
| Accepted domain set to Authoritative, all users in Exchange Online | 250 for real users, 550 5.4.1 for unknown | DBEB is in effect |
| Accepted domain set to Internal relay | 250 for everything | Microsoft advises leaving the domain as internal relay until all valid recipients are in Exchange Online, so unknown addresses are passed on rather than rejected |
| Hybrid with on-premises Exchange | Often 250 for everything | Some recipients exist only on-premises |
| A gateway in front of Microsoft 365 | Whatever the gateway does | The MX points to the gateway, not to Microsoft |
| Unfamiliar or poorly configured verifier IP | 4xx deferral or 5.7.x rejection | Throttling or policy, reported as unknown |
Microsoft also notes "infrequent instances" where non-existent addresses are allowed through even on authoritative domains. So a 250 from Microsoft 365 is strong evidence, not proof.
On the consumer side, Microsoft's Outlook.com policies ask senders not to use namespace mining. Repeated recipient probing from one IP is exactly what that looks like, and is a fast route to having the IP throttled.
How Google Workspace responds to RCPT
Google Workspace normally rejects unknown recipients during the SMTP conversation with a 550 5.1.1 reply saying the account does not exist. A random-address control probe will therefore be rejected, and a 250 for a real address is meaningful.
Workspace domains look catch-all when the administrator has chosen to collect misaddressed mail. Google documents this in get misaddressed email in a catch-all mailbox: a routing rule for inbound messages that changes the envelope recipient to a chosen mailbox or group, applied to "All inactive and unrecognized accounts". With that rule in place every address on the domain is accepted.
This is a deliberate business choice, common in small firms that do not want to miss enquiries sent to a guessed address. Mail to unknown addresses on such a domain goes to a shared inbox, which someone may or may not read.
Google, like Microsoft, also defers or refuses connections from IPs it does not trust. A verifier on a residential IP or without reverse DNS gets unknowns, not answers.
What gateways change
Many organisations put a secure email gateway in front of Microsoft 365 or Google Workspace. The domain's MX records then point to the gateway, so the gateway, not the mailbox platform, answers the verifier.
$ dig +short MX kestrel-holdings.example
10 kestrel-holdings-example.mail.protection.outlook.com. <- Microsoft 365 direct
$ dig +short MX brightside-group.example
10 aspmx.l.google.com. <- Google Workspace direct
$ dig +short MX riverside-legal.example
10 eu-smtp-inbound-1.mimecast.com. <- gateway (Mimecast)
$ dig +short MX oakline-partners.example
10 mx0a-00123456.pphosted.com. <- gateway (Proofpoint)
Whether a gateway checks recipients at the RCPT stage depends on how it is configured. A gateway synchronised with the organisation's directory can reject unknown recipients just as DBEB does. A gateway set to accept everything and filter afterwards answers 250 to every address, including the random control, so the domain is labelled catch-all. From outside you cannot see the configuration; you only see the answers.
RFC 5321 permits this design. It notes that some servers "do not perform recipient verification until after the message text is received", and treat failures as later bounces. The cost of that design falls on the sender: a "valid-looking" address can bounce after you send.
There is no published data on how many gateways accept all recipients, so be wary of any article that gives a percentage.
Signals that still help
When the SMTP answer is uninformative, other evidence carries the decision:
| Signal | How to get it | What it tells you |
|---|---|---|
| Where the address was found | Your extraction source | Published on the company's own contact page is the strongest signal on a catch-all domain |
| Inbox purpose | The local part and page context | hello@ and sales@ on a contact page are meant to receive mail |
| The MX host | dig MX | Whether a Microsoft or Google answer is direct or via a gateway; direct answers are more trustworthy |
| A clean DBEB or Workspace rejection of the control address | The catch-all probe | The domain is not catch-all, so the 250 for your address stands |
| Pattern evidence | Two or more published named addresses on the same domain | Raises confidence in a guessed named address, but does not confirm it |
| Previous sends | Your own bounce history | A delivered message last month is better evidence than any probe |
| Your own suppression list | Past bounces and opt-outs | Never re-probe or re-send to an address you already know failed |
Things that do not help: re-running the same probe repeatedly, switching IPs to get a different answer, or trusting a vendor's "valid" label on a catch-all domain without knowing how it was produced.
Send policy for these domains
| Domain behaviour | Address source | Decision |
|---|---|---|
Microsoft 365 or Workspace direct, 250 and control rejected | Any | Deliverable; send |
Microsoft 365 or Workspace direct, 550 5.4.1 or 550 5.1.1 | Any | Invalid; suppress |
| Catch-all (gateway, internal relay or routing rule) | Published on company site | Send in a capped catch-all segment |
| Catch-all | Pattern guess with published examples | Hold; send only in very small batches if at all |
| Catch-all | Pattern guess with no published examples | Skip |
| Unknown (deferred or refused) | Any | Re-check later from a reputable network; otherwise treat as catch-all |
For the catch-all segment, keep it separate so its bounces are visible, cap it at a small share of each day's sends, and stop it first if the campaign's bounce rate rises. How catch-all sits next to the other verifier results is in email verification results explained.
Remember that a deliverable verdict is not permission. Under the ICO's B2B guidance, you can email corporate subscribers without consent under PECR if you identify yourself and offer an opt-out, but named addresses such as j.patel@ are personal data under UK GDPR and need a lawful basis. Other countries differ.
How Map Extract helps
The email verifier in Map Extract runs an optional SMTP mailbox probe on port 25 with a random-recipient catch-all control, so a Microsoft 365 or Workspace domain that rejects the control is reported as deliverable or invalid, and one that accepts it is reported as catch_all. Where the server's answer is not specific, it records accepted_unconfirmed; where the conversation fails, unknown with a reason such as mailbox_inconclusive or dns_unavailable. The reason is stored with the verdict.
The email extractor records where each address was found and its inbox purpose, which is the evidence that matters most on catch-all domains. In the bulk email sender, settings allow or block catch-all and unknown results, and the policy is checked again at send time.
What it does not do: it cannot see a gateway's configuration, it cannot turn a catch-all domain into confirmed mailboxes, and it does not rotate IPs or retry aggressively to force an answer. Many home and office networks block outbound port 25, which produces unknown results.
Related guides
Key takeaways
- Microsoft 365 with DBEB and Google Workspace normally reject unknown recipients with
550 5.4.1or550 5.1.1. - Catch-all results come from gateways that accept everything, internal relay or hybrid setups, and deliberate catch-all routing.
- Unknown results on these domains usually mean your verifier was throttled, not that the address is bad.
- The MX record tells you who answered; the address's source tells you most about whether it is real.
- Segment catch-all addresses, cap them and stop them first when bounces rise.
Frequently asked questions
Is Microsoft 365 a catch-all email domain?
Not by default. When a Microsoft 365 tenant's domain is set as authoritative, Directory-Based Edge Blocking rejects unknown recipients at the perimeter with 550 5.4.1 Recipient address rejected: Access denied, which gives a verifier a clear answer. Domains look catch-all when they are set to internal relay during a migration or hybrid setup, or when a security gateway in front of Microsoft 365 accepts everything.
Does Google Workspace accept all email addresses?
Not normally. Google Workspace rejects mail for addresses that do not exist unless the administrator has added a routing rule that sends inactive and unrecognised accounts to a catch-all mailbox. Some organisations do this deliberately so misaddressed mail is not lost. When they do, every address on the domain looks valid to a verifier.
Why does my email verifier say unknown for Office 365 addresses?
Usually the verifier's connection was deferred or refused, not the address. Microsoft and other large providers throttle or reject connections from unfamiliar IP addresses, especially those without proper reverse DNS or with a history of probing. The verifier then reports unknown. Re-check later from a reputable network, or rely on other evidence.
How can I tell if a domain uses Microsoft 365, Google Workspace or a gateway?
Look up its MX records. Microsoft 365 domains usually point to a host ending in mail.protection.outlook.com, Google Workspace domains to Google's aspmx hosts, and domains behind a security gateway to the gateway's own hostnames, such as hosts ending in mimecast.com or pphosted.com for Proofpoint. The MX tells you who answered the verifier.
Sources
Written by the Map Extract editorial team. Found an error? Email support@mapextract.uk and we will check and correct it. How we research guides: editorial policy.