Email continuity / phased cutover

Email coexistence during a migration

Email coexistence is an agreed routing design, not a generic claim of zero-downtime migration: test both directions and every business-critical sender.

In a phased migration some employees use the old mail platform while others use Microsoft 365. A company may have correctly copied mailbox history but still lose new messages through conflicting accepted domains, a forgotten application sender or a connector that forwards mail the wrong way. The migration plan must therefore state how new external mail is routed and how replies, aliases and business applications behave while both systems exist.

The safe order is to document the current mail flow, configure a supported coexistence or staged-routing model, test it with representative real addresses, and only then make a public DNS or accepted-domain change. A successful MX lookup by itself proves neither end-user receipt nor authentication alignment.

Discuss an actual project →

Workstream 01

Start with a sender, recipient and domain inventory

List every verified domain, secondary SMTP alias, shared mailbox, distribution list, forwarding address and application that sends business email. Include CRM notifications, invoice systems, marketing platforms, web forms and printers. A platform migration may alter where an address exists without migrating the sender that uses it.

Trace at least one example of external-inbound mail, external-outbound mail, employee-to-employee mail, delegated replies and a system-generated message. Record the message path before changing anything so that aftercare can distinguish a new failure from a pre-existing one.

Flow to proveWhat to testTypical hidden dependency
External inboundSender outside both tenants to moved and unmoved usersMX and recipient routing
External outboundReply and new message from both environmentsSPF/DKIM/DMARC alignment
Shared identitySend-as/delegation from approved employeeMailbox rights and X500 history
Application mailCRM or website notification arrives at outside recipientSMTP credentials and third-party DNS
Calendar invitationsFree/busy and update to external inviteRecipient resolution and coexistence

Workstream 02

Choose the correct coexistence model for the business

Microsoft supports a range of Exchange Online mail-flow and connector options depending on whether the source is another tenant, an on-premises Exchange estate or an external hosted system. The design must reflect who owns the accepted domain at each stage, which platform is authoritative for recipient discovery, and where undeliverable mail is reported.

A cutover can involve temporary domains, mail contacts or routing addresses, transport connectors and staged MX changes. Avoid proposing a generic forwarding chain without considering loops, spam filtering, service security and the effect on replies. A clear diagram showing current and transitional paths is a useful deliverable.

What to capture before the decision

  • Confirmed control of each verified domain and registrar
  • Approved transitional address/routing design
  • Known source and target recipient inventory
  • Time window and authority for DNS/connector edits
  • Documented test for any transport loop or ambiguous alias

Workstream 03

Make authentication changes as carefully as MX changes

Every sender using the domain must be considered when SPF, DKIM and DMARC are reviewed. An MX update changes where mail is received; it does not automatically authorise outbound mail from Microsoft 365 or remove an old vendor's sending authority. A premature SPF replacement can cause invoices or contact-form messages to fail authentication.

Review the existing TXT, CNAME and MX configuration, then plan a controlled change with propagation allowances and a validation procedure. Microsoft documentation explains how SPF, DKIM and DMARC work together; approval depends on the actual combination of business senders.

Workstream 04

Acceptance and recovery after the final routing change

Use independent senders and recipients outside both tenants to prove delivery and reply routing. Exchange message trace can show whether mail was delivered, rejected or deferred; combine that technical evidence with the end user's inbox confirmation. Test priority shared mailboxes and non-human senders as separate cases.

Identify the people authorised to pause the cutover or reverse DNS changes where technically feasible. A DNS rollback does not automatically undo mailbox identity changes or a completed native migration; an agreed recovery procedure must reflect the state of each workload. Retain logs, an exception register and user communications.

What to capture before the decision

  • External end-to-end mail round trips passed
  • All expected aliases and shared mailboxes tested
  • Application senders confirmed with their owners
  • Authentication and message-trace evidence retained
  • Final routing owner signs off the operating state

Coexistence succeeds when ordinary correspondence keeps working during the transition and the business can explain exactly where every important address is delivered.

Source documents and limitations

These references establish current vendor behaviour; they do not verify an individual customer's configuration or imply Microsoft affiliation. Technical claims should be rechecked before execution.

Source and destination first

Review the email coexistence and cutover plan

Tell us your source and destination, domains, user waves and any critical application senders. We can help define measurable routing and acceptance conditions.

Discuss scope