SyncTrix logoSyncTrix
All articles
Cloud11 min read

Microsoft 365 after the July 2026 rise: when do Australian subsidiaries need separate mail domains to stop paying for dormant seats?

If multiple Australian entities share one Microsoft 365 mail design, you may be carrying avoidable seat cost. Here is when separate mail domains help, and when they do not.

By SyncTrix Editorial
Microsoft 365 after the July 2026 rise: when do Australian subsidiaries need separate mail domains to stop paying for dormant seats?

Many Microsoft 365 cost reviews start with plan comparison and stop there. That misses a common source of waste in multi-entity groups: the way email domains, recipients and tenancies were set up when the organisation was smaller or less structured. After Microsoft's pricing update took effect on 1 July 2026, with existing customers moving to the new price at renewal, small pockets of licence waste became easier to see because the baseline per-user cost moved up across key commercial plans. Microsoft lists Office 365 E3 at USD 26 per user per month, up from USD 23, Microsoft 365 E3 at USD 39, up from USD 36, Microsoft 365 F3 at USD 10, up from USD 8, and Microsoft 365 Business Basic at USD 7, up from USD 6, noting that local pricing varies by country and currency (Microsoft licensing update). If your Australian group still keeps former staff, parked entities, internal-only users or lightly active subsidiaries inside one shared mail namespace, the technical question is no longer just 'what licence do they need'. It is 'does the current domain and tenancy design force us to keep licensed user mailboxes where a different structure would let us keep addresses, routing and retention without paying for full seats?' This post looks at when separate mail domains are the right answer, when they are not, and what to check before you split anything.

01Start with the real problem: address continuity is being confused with mailbox licensing

In many groups, the parent company kept one primary mail domain for everyone and then added subsidiaries into the same tenant as headcount grew. That is technically easy because Exchange Online lets you add a domain as an accepted domain so recipients in that domain can send and receive mail inside the tenant (accepted domains). The problem comes later, when entity boundaries, HR processes and cost allocation diverge but the mail design does not.

An Australian subsidiary is a separate legal entity with its own ABN when incorporated and carrying on an enterprise, even if it sits inside the same corporate group (ABR guidance). That legal separation does not force a separate Microsoft 365 tenant or domain. But it often changes the business case. If a dormant or low-activity entity still needs its brand address to exist, that does not automatically mean every historical or role-based mailbox under that entity still needs a paid user licence.

This is where many renewals go wrong. Teams ask, 'Can we delete the mailbox?' when the real question is, 'What exactly must survive: the address, the mail history, the ability to receive, the ability to send, or the legal hold?' Once you separate those requirements, you can usually see which seats are genuinely unavoidable and which are artefacts of earlier design. If you are already reviewing other Microsoft 365 clean-up items, the same discipline applies to ex-staff retention choices and resource and service-account exits.

  • Keep the address, no user works there now: often a shared mailbox, contact, or forwarding design issue, not a user licensing issue
  • Keep historic email for compliance: often an inactive mailbox issue
  • Keep a live mailbox because someone still works there: likely a real seat
  • Keep a whole entity inside the parent namespace only because changing addresses feels disruptive: that is a design inertia issue

02What Microsoft 365 actually lets you do without keeping a paid user seat

Microsoft's own documentation is clear on two important options. First, a shared mailbox can hold up to 50 GB without its own licence. It is intended for addresses like support, reception and other mailboxes used by multiple people. The catch is that each person accessing it must already have a licensed Exchange Online mailbox, and if the shared mailbox needs more than 50 GB, archiving or litigation hold, extra licensing applies (shared mailbox licensing).

Second, inactive mailboxes let you retain a former employee's mailbox after deleting the user account, provided you correctly apply a hold first. Microsoft states that the mailbox must be licensed correctly so the hold can be applied before deletion, and once the account is deleted the Exchange Online licence becomes available for reassignment (inactive mailbox process). Microsoft also states that inactive mailboxes cannot receive email and are not shown in the address book, so they solve retention, not ongoing address continuity (inactive mailbox behaviour).

Those two facts matter because they show that many 'dormant seats' are not dormant at all. They are mixed-purpose mailboxes doing at least two jobs at once. A former employee mailbox is being kept for legal retention and as a live contact point. A small subsidiary mailbox is being kept as a brand address and as a historic archive. A departed office manager's mailbox is still receiving supplier mail because nobody moved the public contact address. If you keep both purposes bound to one licensed user mailbox, the seat looks necessary even when it is not.

There is also a compliance nuance. Microsoft's licensing FAQ says customers that have Exchange Plan 2 through standalone licensing or Microsoft 365 or Office 365 E3 or E5 can set mailboxes as inactive by applying a hold, which frees the licence to be reassigned while preserving mailbox data (licensing FAQ PDF).

03When separate mail domains help, and when they are a distraction

A separate mail domain for each subsidiary helps when the domain boundary matches an operating boundary. Typical examples are entities with their own customer-facing brand, separate staff population, different joiner and leaver patterns, or a realistic chance that the entity will be sold, wound down or moved to another tenant later. In those cases, keeping the entity on its own domain, such as entityname.com.au, makes it easier to convert people, aliases and role addresses into the right recipient types without preserving licences just to keep addresses under the parent's main namespace.

By contrast, separate domains do not solve waste if the actual issue is poor recipient hygiene inside one entity. If twenty ex-staff mailboxes are still licensed because finance occasionally searches their old correspondence, splitting the domain changes nothing. Those should be assessed as inactive mailbox, shared mailbox or forwarding cases first. Likewise, if the subsidiary has active staff who need full Outlook, OneDrive and Teams, changing the domain does not change the need for seats.

Exchange Online accepted domains support multiple domains in one tenant, and Microsoft distinguishes between authoritative and internal relay domains. Authoritative domains reject mail for unknown recipients. Internal relay domains can relay mail for recipients that are not known inside Microsoft 365, provided the right connector path exists (accepted domain types). That means you do not need a separate tenant just because you want a separate domain. Often, the practical move is to keep one tenant but give each subsidiary a cleaner domain boundary and recipient policy.

The bigger question is whether the subsidiary's current dependence on the parent's primary domain is what forces you to keep mailboxes alive. If incoming supplier, customer or recruitment mail still lands on named user addresses in the parent namespace, then yes, a separate subsidiary domain can be the key that lets you retire those named mailboxes and replace them with role-based mailboxes or contacts. If the subsidiary already operates mostly through role addresses, separate domains may give you tidier governance but little direct licence saving.

  • Helpful: separate brand, separate operating model, likely future divestment, repeated ex-staff licence retention under a shared namespace
  • Not helpful by itself: active users still need full productivity features, retention is the main issue, role addresses already exist, nobody plans to move entity boundaries
QuestionIf yesIf no
Does the entity trade under a distinct external brand?A separate domain probably reduces future address clean-upDomain split is less urgent
Are former staff mailboxes being kept licensed mainly to preserve the address?A domain and role-address redesign may cut seatsLook first at retention and mailbox type
Would the entity plausibly move to another tenant later?Separate domain now reduces migration frictionOne-tenant, multi-domain may be enough
Do customers and suppliers still write to named people rather than role addresses?Redesign addressing before renewal if possibleDomain separation may bring limited savings
Does the mailbox need ongoing mail receipt, not just retention?Inactive mailbox is not enoughInactive mailbox may release the seat
Quick test: does this subsidiary need a separate mail domain now?

04The Australian subsidiary pattern where seat waste usually appears

The common pattern is not one giant licensing error. It is a cluster of small entity-specific exceptions. A newly quiet subsidiary still has a handful of named mailboxes because some suppliers never updated contacts. Another entity had a manager leave, but their mailbox stayed licensed because legal wanted access to historical mail and operations wanted messages to keep arriving. A local brand stopped hiring, but jobs@ and accounts@ still point into individual user mailboxes. None of those cases is unusual. Together, they create a tail of paid seats that survive every annual clean-up.

This is also where domain design intersects with corporate structure. Because the subsidiary is a separate entity in Australian legal terms, it often makes sense for that entity's persistent addresses to be entity-level assets, not person-level mailboxes. If those assets live under their own domain or at least their own clearly bounded accepted domain inside the tenant, you can separate role addresses from individuals earlier. That makes leavers easier to de-license.

If you are thinking about a broader redesign, keep the blast radius small. In most groups, you do not need to move every user on day one. Start with one subsidiary that has either dormant operations, low headcount, or a high ratio of named historical mailboxes to current staff. The aim is to prove that a cleaner recipient model changes the licence baseline at renewal. If the entity later needs its own tenant because of M&A, regional policy or operating independence, Microsoft does support cross-tenant mailbox migration, but Microsoft also notes that cross-tenant moves require a per-user one-time migration licence and that mailboxes on hold are blocked from migration (cross-tenant migration). That is another reason to fix your mailbox categories before a separation event, not during it.

05How to audit the issue before your next renewal

Do not start with licence SKUs. Start with recipient inventory. Export every mailbox, shared mailbox, mail user, contact, distribution group and accepted domain for the tenant. Then tag each recipient with five business fields: entity owner, still staffed, needs to receive mail, needs to send mail, and retention requirement. That one exercise usually shows whether the current design is hiding paid seats.

Next, group recipients by domain and entity, then look for named user mailboxes that fail a simple test: no active staff member, but the mailbox is still licensed. Those are your first candidates. Some will convert to inactive mailboxes if retention is the only need. Some will convert to shared mailboxes if a team still works the inbox and it fits Microsoft's shared mailbox conditions. Some will become mail contacts or forwarding targets. A smaller number will remain real user seats.

After that, inspect namespace design. Which subsidiaries still rely on the parent domain as their public address identity? Which role addresses are aliases on top of user mailboxes instead of separate recipients? Which subsidiaries have future sale, closure or tenant separation risk? Those are the entities where separate mail domains have financial value, because they let you stop treating address persistence as a reason to keep user objects licensed.

The billing angle matters too. Microsoft's new commerce model supports monthly, annual or three-year terms for many subscriptions and lets changes take effect at end of term in planned ways (new commerce overview). If your estate is on annual commitments, the economic benefit of fixing bad recipient design before renewal is far higher than discovering it just after the renewal date. If you need help sequencing the technical work, this usually sits between mail infrastructure, cloud infrastructure and governance and compliance rather than being just a licensing spreadsheet exercise.

  • Export and classify recipients before analysing licences
  • Separate retention needs from live address needs
  • Find named mailboxes owned by entities, not people
  • Map each subsidiary to a clear public mail identity
  • Time changes to renewal and commitment dates

06Questions to ask your Microsoft partner or mail vendor

A good vendor should be able to answer structure questions, not only SKU questions. Ask them to quantify how many currently licensed user mailboxes exist only because the organisation wants to preserve an address, retain historical mail, or keep a quiet subsidiary reachable. Ask them which of those could move to shared mailbox, inactive mailbox, contact or relay patterns under Microsoft's published rules.

Ask whether the proposed design keeps one tenant with multiple accepted domains, or introduces a second tenant. One tenant with multiple domains is usually the simpler first step. If they push straight to a second tenant, ask what future operating constraint makes that necessary now. Also ask what happens if a subsidiary is sold later. Can they show a clean path for cross-tenant migration and flag the impact of holds and one-time migration licensing from Microsoft's own documentation.

Finally, ask them for the effort estimate in engineering terms, not just consultancy language. You want counts of domains, recipient types, transport rules, connectors, identity updates, user communication steps and rollback points. If they cannot explain how mail flow behaves for authoritative versus internal relay domains, or when a connector is required, they are not really designing the operating model (accepted domain routing).

  • How many paid user mailboxes are currently preserving addresses rather than serving active staff?
  • Which of those can move to shared or inactive mailbox patterns under Microsoft rules?
  • Which subsidiaries should get separate accepted domains inside the current tenant?
  • Do any domains need internal relay and connectors, and why?
  • What future divestment or tenant split scenarios does this design make easier or harder?
  • What exact recipient and mail-flow changes will be made before renewal?

07What a sensible decision looks like

If an Australian group has multiple subsidiaries but one stable operating model, one tenant with multiple accepted domains is often enough. The win comes from moving entity contact addresses away from named licensed users and using the right mailbox type for the right purpose. Separate domains help when they create a durable boundary between entity identity and individual employment.

If a subsidiary is largely dormant, has external addresses that must remain live, and still consumes user licences mainly to keep those addresses reachable, then yes, domain redesign is often the trigger that exposes and removes avoidable seat cost. If the subsidiary still has active knowledge workers using the full Microsoft stack every day, separate domains may improve governance and future flexibility, but they will not materially reduce seat spend on their own.

The practical rule is simple: do not redesign domains because Microsoft raised prices. Redesign them when the current namespace forces you to buy user seats for non-user reasons. That is the point where architecture becomes a licensing lever. Teams like SyncTrix typically handle this as a mail operating model review first, then a controlled recipient and routing change, rather than as a tenant-wide migration project.

Topics

Microsoft 365 separate mail domains AustraliaMicrosoft 365 dormant seats subsidiariesExchange Online accepted domains multi entityAustralian subsidiary Microsoft 365 tenancy designinactive mailbox shared mailbox licensingMicrosoft 365 renewal July 2026 Australia

SyncTrix Editorial

SyncTrix · SyncTrix

Writes about the engineering decisions behind production systems - architecture, delivery and the trade-offs that only show up at scale.

Building something like this?

SyncTrix engineers AI, SaaS, platform and cloud systems for enterprises and high-growth teams. Tell us what you're shipping and we'll scope it with you.