One tenant after the deal closed, without a cutover weekend gamble

A mid-market professional services firm closed an acquisition and immediately felt the drag of two Microsoft 365 tenants: separate identities, awkward guest collaboration, and leadership asking when everyone would “just be on one system.” They needed a single surviving tenant, without betting payroll week on a single big-bang cutover.

• ~1,200Users consolidated
• 8 weeksDiscovery through decommission
• 6 wavesPilot + departmental cutovers
• 0Unplanned all-hands outages

The situation

The acquiring company already ran a mature Microsoft 365 tenant. The acquired firm had its own tenant with roughly 1,200 mailboxes, years of SharePoint growth that nobody had cataloged, and Teams heavy on guest access from client engagements.

Legal wanted one primary collaboration plane. Finance wanted license sprawl to stop. End users wanted their email addresses to keep working Monday morning. IT had lived through a painful migration years earlier and was rightly skeptical of “we’ll do it over the weekend” plans.

What made it hard

• Two brand domains that both needed to keep receiving mail during the project
• Guest-heavy Teams used for client work, easy to break if membership isn’t planned
• Uneven SharePoint hygiene, active project libraries mixed with abandoned sites
• App consents and SSO tied to the source tenant that finance and delivery tools still used
• Executive calendars and shared mailboxes that couldn’t tolerate long dual-identity confusion

How we approached it

1. Discovery before dates

We started with exports and working sessions, not a go-live calendar. Inventory covered users and guests, mailbox sizes and holds, SharePoint/OneDrive usage, Teams and private channels, Conditional Access, enterprise apps, and every verified domain.

That work produced a plain-language risk list leadership could actually prioritize: what had to move, what could wait, and what would deliberately not migrate.

2. Identity and domain strategy first

We designed UPN and SMTP mapping so primary addresses stayed familiar. Domain moves were sequenced so neither tenant went dark mid-project. Coexistence covered free/busy where it mattered for the integrated project teams.

3. Target tenant prepared like production, not a sandbox

Before the first production wave, the surviving tenant had MFA expectations, Conditional Access baselines, licensing headroom, and Teams/SharePoint policies that matched how the combined firm actually worked, not a default Microsoft setup.

4. Pilot that looked like real work

Twenty-five pilot users spanned IT, finance, a delivery pod, and two executives. Validation wasn’t “mail arrived.” It was Outlook mobile, shared calendars, OneDrive files, critical Teams channels, and the two SSO apps people open every morning.

Defects went into the runbook. We didn’t expand waves until pilot sign-off was written down, not assumed.

5. Departmental waves with short cutover windows

Production moved in six waves. Each wave used staged sync, a cutover checklist, and a known support channel. Communications told people which tenant to sign into and when. Delta sync kept finalization windows short instead of multi-hour “copy everything again” nights.

6. Cleanup and decommission

After waves completed, we cleaned orphaned permissions, reclaimed licenses, exported what retention required, removed domains cleanly, and shut down the source tenant with a checklist finance and IT both signed.

Timeline at a glance

Results

• About 1,200 users landed in the surviving tenant without an all-hands mail outage
• Integrated delivery teams collaborated in one Teams environment within the first production waves
• Both brands’ primary SMTP addresses stayed usable through the coexistence window
• Source tenant retired after validation, no lingering dual-license surprise for finance
• IT kept a reusable wave playbook for any future tuck-in acquisitions

We needed one tenant without betting the business on a single cutover weekend. The wave plan and validation checklist kept us in control.

- Director of Technology, professional services firm

What we’d tell a peer in the same spot

Don’t start with a tool license and a go-live date. Start with identity, domains, and a pilot that mirrors real work. The migration platform matters, but the runbook, communications, and security parity in the target tenant decide whether people trust the project after week one.