Home › Services › Tenant consolidation
Tenant consolidation after M&A, or a carve-out on divestiture
One accountable plan for bringing two or more Microsoft 365 tenants together — or carving one out. Dual-tenant discovery, target architecture, licensing reconciliation, a wave plan with rollback gates, and day-one coexistence so both businesses keep working. Not fifteen disconnected migrations with nobody holding the schedule.
There is no merge button. Two Microsoft 365 tenants cannot be joined, and a division cannot be split out, by any switch Microsoft provides. Every consolidation decomposes into a sequence of moves — identities, mailboxes, archives, SharePoint sites, Teams, OneDrive, domains, sometimes Azure — each with its own tooling, its own limits and its own way of failing.
The individual moves are the part we have done most often. What consolidations actually fail on is everything between them: putting them in an order that works, keeping both businesses operating while the waves run, reconciling two licensing agreements before renewal dates lock them in, and holding a dozen workstreams to one schedule when the deal date moves.
That coordination is what this engagement is. It runs in both directions: acquirers use it to collapse acquired tenants into one, sellers use it to carve a division into a clean tenant before a transaction closes, usually against a transitional service agreement deadline.
We have run it at every size that matters. Seventeen tenants into one for a healthcare group in Australia. Eight into three for a group in the UAE and Singapore, then one carved back out eighteen months later when they divested. Three thousand users and 24 TB across a European relocation, with no planned downtime. The write-ups, including the project where the migration host died mid-flight, are published.
The success criteria we hold ourselves to, and what the closing report has to show.
Documents, not slides. Each one exists because a project went wrong without it.
Weeks are indicative for a two-tenant consolidation of moderate size. Your quote states the schedule for your deal.
Read-only assessment of every tenant in scope. Nothing changes. Users, domains, workloads, data volumes, licensing, security posture, third-party dependencies. Duplicate licensing surfaces here, before anything moves, which is often the first thing that pays for the engagement.
Identity design, naming conventions, domain strategy and the licensing model for the combined or carved-out organisation. This is where the arguments happen, on paper. It is also where we tell you which parts of the schedule are fixed by the platform and not by us.
Each workload move is scoped, then assembled into a wave plan with dependencies and rollback gates. Domain timing drives much of it, because a custom domain can only be attached to one tenant at a time.
How mail, calendars, chat and files behave across the tenant boundary while the waves run. Designed before the first wave, not discovered during it.
Identities first, then data, then decommission. Large archives start on day one and run in the background regardless of where the rest of the plan has reached — they finish when they finish. Each wave ends with its own validation.
Domain moves, final identity switches, and the communications that go with them. Executed from the runbook, with rollback gates honoured rather than skipped to hit a date.
Target-state verification, source decommission checklist, licence cleanup against renewal dates, and the closing report with the full paper trail.
Send us the tenant count, the rough headcount and what is forcing the date — a close, a TSA expiry, or nothing yet. We come back with the sequence, an honest view of what will be slow, and a fixed price.
Fixed, after scoping · 10–14 weeksSplit out explicitly, because the two most common causes of delay are an unsigned decision and an assumption about who owned it.
The constraints that are fixed by the platform rather than by us, and the ones we learned the expensive way.
Both, with our own engineers. The programme covers assessment, architecture, sequencing, coexistence, communications and orchestration; the workload moves are executed by the same team. There is no handover to a subcontractor halfway through, and no third party seeing your data that you have not been told about.
Ten to fourteen weeks is typical for a two-tenant consolidation of moderate size: roughly two weeks of discovery, two of architecture and sequencing, and the rest execution and closure. Under 100 users can run shorter. A thousand or more usually runs six months or longer. The honest variables are data volume, tenant count, TSA dates and how quickly decisions get signed.
That is the design goal of the coexistence plan, and it is what we have delivered. Mail flow, calendaring, chat and file access across the tenant boundary are designed before the first wave runs. Our consolidations have not required planned downtime, including a 3,000-user relocation moving 24 TB.
A fixed price, quoted after we have looked at both tenants. Scoping costs you nothing, and the calculator will give you a first number before you speak to anyone. The price is driven by user count, how much data actually needs to move, and how tangled the identities are — not by how long the project ends up taking us. No hourly billing and no variation orders.
Yes — a carve-out is the same discipline run in reverse. Discovery establishes what belongs to the divested entity, a clean target tenant is designed and stood up, and the waves move that division out. We have done both for the same client group: eight tenants into three, then one carved back out eighteen months later. Sellers usually run this against a TSA expiry rather than a convenience date.
If the domain moves with them, yes. But a custom domain can be attached to only one tenant at a time, so the domain move is a scheduled cutover event rather than a gradual transition. Where an acquired brand is being retired, users typically take the acquirer’s domain with the old address preserved as an alias. The domain plan is decided in the architecture phase, before any wave runs.
They are mapped and decided case by case. Some merge into a single identity; some stay separate because they genuinely hold different roles in different parts of the business. On both of our largest consolidations, more than half were deliberately kept separate. Merging identities that belong to different roles creates more problems than it solves.
No, and usually you should not. Archived mail nobody has opened in years is the largest single driver of migration time. We push hard at kick-off to agree what has to move, what can be archived elsewhere and what can be left behind under a retention policy. It is an uncomfortable conversation at kick-off and a far worse one in month five.
TSA expiry is treated as a fixed constraint from day one: the plan is built backwards from it and the riskiest dependencies are scheduled earliest, so slippage is visible in week two rather than week seven. What we will not do is compress validation to hit a date. If the deadline and the scope genuinely conflict, you hear it during sequencing while there is still time to renegotiate.
You do, entirely. We act as Cloud Solution Provider of record for the subscriptions only if you want us to, and that can be changed at any time. Nothing about the consolidation locks you to us.
Whichever of ours is authorised for your country, in the currency Microsoft sets for it — euros, sterling, francs, kronor, kroner, dirhams, Australian or Canadian dollars. For a group with entities in several countries, each gets a local invoice under one relationship. The United States is served by our headquarters location at o365hq.com.
Four things, and we will tell you in week two if any of them apply: personal archives above 100 GB, an unsigned target architecture, a works council approval that was not started in parallel, and a third-party application that authenticates against the source tenant and nobody owns. None of them are exotic. All of them are visible during discovery if someone looks.
Scoping costs nothing
How many tenants, roughly how many users, and what is forcing the timing — a close date, a TSA expiry, or nothing yet. You get back a sequence, an honest view of what will be slow, and a fixed price. Usually within one business day, and nobody rings you unless you ask.