Every Office 365 conversation in early 2016 starts the same way: someone saw a slide about cloud mail and asked when we can “just move.” We start somewhere less glamorous — Exchange 2010 capacity and growth truth. If you do not know what the estate is doing on the floor, you cannot size hybrid, network, or risk. This post is how we inventory an Exchange 2010 environment before anyone opens the Migration Advisor.
What we pull first
- Mailbox count, archive count, and disconnected mailboxes
- Average and P95 mailbox size; top 50 outliers
- Database count, sizes, and backup window duration
- DAG membership, lagged copies, activation preference
- Transport servers, edge, journaling, third-party archiving
- Client mix: Outlook versions, ActiveSync partnerships, OWA usage
Get-MailboxStatistics and database reports are not optional homework. They are the business case inputs.
Growth is the silent killer
A farm that “fits” today may not fit the 2017 hardware refresh if growth stays on-prem. We chart twelve months of database growth where logs allow. Finance understands “we buy another SAN shelf in 18 months or we move mail.” They do not understand vague cloud enthusiasm.
DAG and ops cost
Exchange 2010 HA is real work: patching coordination, failover drills nobody runs, certificate cycles, storage alignment. When we price “stay on-prem three more years,” we include people weekends, not only disk. That is often where Office 365 starts to look rational even before feature debates.
Network is a later slide — not never
Egress math for hybrid and cutover comes after capacity. But we already note whether the internet edge is a single firewall pair shared with everything else. That foreshadows assessment work we will do mid-year on larger programs.
Deliverable
A short capacity brief: current state, 24-month on-prem trajectory, risks (backup window, single points), and a recommendation — stabilize, expand, or open a formal Office 365 assessment. No tool licenses required for this step. Honesty is the tool.
Mailbox outliers and the politics of size
Every estate has fifty mailboxes that skew averages: executives with decade-old PSTs ingested as archives, departments that never deleted, shared mailboxes used as file shares. We isolate outliers into their own workstream. They need archive policy, legal hold review, or staged migration — not a single wave with everyone else. Ignoring them produces “the migration is slow” narratives that are really “we tried to move a file server disguised as mail.”
We also separate user mailboxes, shared/resource mailboxes, and system mailboxes. Counts that mix them make staffing estimates wrong by double-digit percentages.
Backup windows as a capacity signal
If full backups routinely spill past the maintenance window, you already have a capacity crisis — storage growth, IOPS, or both. Cloud migration does not erase backup; it changes who owns retention and how eDiscovery works. We record current backup duration and failure rate so finance sees operational pain, not only disk SKUs.
Client access patterns
Outlook Anywhere, ActiveSync device counts, and OWA usage tell you who will feel hybrid coexistence pain. A mobile-heavy field force has different cutover risks than a desktop-bound campus. We sample ActiveSync partnerships and flag ancient devices that will fight modern auth later — even in 2016 that debt is forming.
What a good capacity brief contains
- Current counts and twelve-month growth trend
- Top storage consumers and proposed treatment
- HA topology risks (single DAG copy patterns, lagged copies never activated)
- People cost of remaining on-prem three years (patch weekends, DR tests)
- Decision gate: stabilize, expand capital, or open Office 365 assessment
Sequencing into the rest of the year
Capacity honesty in January feeds tenant readiness and formal assessment mid-year. Skipping it produces assessments full of tool logos and empty of mailbox truth.
Scenario walkthrough
Consider a mid-size organization with hybrid identity, mixed desktop delivery, and a mandate to reduce risk without stopping the business. Week one is inventory and sponsor alignment. Week two is a written target state with two options and explicit out-of-scope items. Weeks three and four are pilot build and measurement. Only then does broad change begin. Compressing that sequence into a single weekend is how outages are born.
Along the way, three conversations dominate: who owns identity decisions, who pays for platform capacity, and what residual risk leadership accepts in writing. When those conversations are avoided, engineers improvise under pressure and the organization inherits accidental architecture.
We keep a living risk register with severity, mitigation, residual risk, and owner. The register is reviewed in the same meeting as the delivery burn-down. Risks that never move owners are the ones that become incidents.
If you're facing this
If leadership wants cloud mail without mailbox statistics, pause and measure. We run Exchange 2010 assessments that feed real migration programs — bring database reports and a year of growth if you have it.