Microsoft Ignite is running in Orlando this week, and Monday's news answered the question every unified-communications architect has been asking since March: Microsoft announced that Teams will become the primary communications client in Office 365, with Skype for Business capabilities — calling, meetings, the works — being built into Teams on a new backend they're calling the intelligent communications platform. Microsoft is replacing Skype for Business with Teams. Not "positioning alongside." Replacing, over time, in the cloud.
We run a multi-tenant Skype for Business Server 2015 platform carrying 10,000-plus concurrent users at 99.95% availability, and we've deployed SfB estates for government and enterprise clients all through 2016 and 2017, so we sat through the sessions with more than academic interest. Back in March, when Teams went GA, we wrote that the incentives pointed exactly here and that our planning assumption was Teams accumulating capabilities toward SfB Online over a couple of years. That assumption is now an announced roadmap. Here's what was actually said, what was carefully not said, and the migration math we're running with clients this week.
What Ignite actually announced
Strip the keynote adjectives and the concrete commitments are these:
- Teams becomes the core communications client for Office 365. Calling and meeting capabilities from Skype for Business are being rebuilt into Teams on a new Skype infrastructure — media stack and all — rather than bolting the old plumbing onto the new client.
- Skype for Business Online continues, but the direction is set. No retirement date was given. Microsoft's own phrasing is that Teams will "eventually" replace SfB Online. Roadmap items — calling features, meeting features, interop — are being published so customers can track parity.
- A new on-premises Skype for Business Server is coming. A vNext server release was committed for the second half of 2018, targeted at customers who can't move communications to the cloud. That sentence matters enormously to our regulated and government clients: on-premises SfB has a next version and therefore a support runway measured in many years.
- Interop and side-by-side. Teams and SfB will coexist — presence, messaging interop, and the ability to run both clients while capabilities converge.
Equally important is what was not announced: no end-of-life date for SfB Online, no parity date for enterprise voice in Teams, no hybrid voice story for Teams yet, and nothing that changes a Skype for Business Server deployment's support status today. Every panicked headline you'll read this week glosses over those four absences.
Reading the announcement like an architect
The phrase worth dwelling on is new infrastructure. Microsoft is not porting the SfB Online backend; they're rebuilding calling and meetings cloud-native. Two consequences follow.
First, parity will arrive feature by feature, and the last 10% will take the longest. IM, presence, and internal meetings are one class of problem. Enterprise voice is another animal entirely: dial plans, response groups, common-area phones, analog devices, contact centers, compliance recording, E911, session border controllers into legacy PBXs. Our voice migrations at scale run 12 to 24 months on a stable target platform. Teams today has no PSTN calling story at all — that's roadmap. Anyone who tells you their 8,000-seat voice estate is moving to Teams next quarter hasn't inventoried their response groups.
Second, the corporate energy follows the strategic platform. SfB Online will be maintained, but the feature curve there is now flat by definition. The practical reading: SfB Online is a stable bridge, not a destination. The destination for cloud UC is Teams; the destination for can't-do-cloud UC is Skype for Business Server vNext in 2018.
The migration math we run with clients
Every client conversation this week lands on the same three variables: workload (chat and meetings versus enterprise voice), residency (cloud-permitted versus on-prem-required), and horizon (how long the current estate must earn its keep). The matrix we're using:
| Your estate | Our read after Ignite |
|---|---|
| SfB Online, IM/presence/meetings only, no voice | You're the early cohort. Pilot Teams for chat now; plan meetings migration as parity lands through 2018. Cheapest move in the room |
| SfB Online with Cloud PBX voice | Hold. Watch the Teams calling roadmap; do not move dial tone to a platform that doesn't have it yet. Revisit quarterly |
| SfB Server 2015 on-premises, meetings + voice driver | Finish and run it. Your platform is supported, a vNext server is committed for 2018, and nothing announced changes your business case. Track parity; plan a 2019–2020 evaluation |
| Mid-deployment on SfB Server right now | Complete the deployment. A 12-month voice project doesn't change direction because of a keynote about a client that can't yet terminate a PSTN call |
| Regulated / sovereignty-bound | Your path is SfB Server vNext. Budget for that upgrade in 2018–2019 and let the cloud story mature on someone else's users |
| Greenfield, no UC today | Genuinely new answer: start chat and internal meetings in Teams now, put voice on SfB (or keep the PBX) as the bridge, and avoid deploying any new SfB footprint you don't strictly need |
Notice the pattern: for exactly one cohort — greenfield — the announcement changes what you should build. For everyone else it changes what you should watch.
There's also an economic line item people miss. Running two clients side by side is not free: two sets of user training, two support scripts at the helpdesk, interop edge cases, and a governance question every time a meeting gets scheduled ("which platform?"). We're advising clients to budget the coexistence period honestly — it's a project phase with real cost, not a passive state you drift through. The tenants that treat coexistence as deliberate (defined cohorts, defined workloads per client, a published end state) will spend a fraction of what the drift-and-hope tenants spend.
What we're doing with our own platform
Our multi-tenant SfB 2015 environment isn't going anywhere. The workloads it exists for — PSTN voice, federated meetings, contractual availability, tenant isolation — are precisely the workloads Teams doesn't address yet, and our clients' contracts run on dial tone, not keynotes. What changes is our roadmap posture:
- We're standing up a Teams evaluation tenant against our own SfB federation to test interop weekly as Microsoft ships it — presence fidelity, message routing, meeting join friction. We'll publish notes as it firms up.
- Every new statement of work now includes a "Teams transition" section: what we'll measure, which parity milestones trigger a migration conversation, and who owns the decision. Writing that section in 2017 is dramatically cheaper than discovering in 2019 that nobody owns it.
- We're telling every voice client the same sentence: your next PBX decision is your last PBX decision. Whatever you sign this year should have an exit that lands on the intelligent communications platform, because by the next refresh cycle, that's where this ends.
# The quarterly homework we've assigned ourselves for every tenant we manage:
# inventory who still lives on SfB and which workloads they use.
Get-CsOnlineUser -ResultSize Unlimited |
Select-Object UserPrincipalName, EnterpriseVoiceEnabled, HostingProvider |
Group-Object EnterpriseVoiceEnabled
Unromantic, but that inventory — who has voice, who only chats, who actually joins meetings — is the entire migration plan in embryo. The tenants that keep it current will migrate in cohorts on their own schedule. The ones that don't will migrate all at once on Microsoft's.
The part nobody at the conference says out loud
Skype for Business isn't dying because it failed. It's being replaced because the center of gravity moved: persistent, channel-based collaboration turned out to be the workload that pulls everything else — and once Teams owned that ground, meetings and calling were always going to follow the conversations. We watched it happen inside our own consultancy this year; project comms moved to Teams without a mandate, while every client call stayed on SfB. Microsoft is now formalizing what usage data already showed them.
That's also the reassurance for anyone feeling burned about a 2016 SfB investment: the estate you built is the bridge, and bridges get you across. The mistake isn't having built it. The mistake would be extending it in 2019 like the announcement never happened.
If you're facing this
If your steering committee saw the Ignite coverage and wants to know whether your Skype for Business investment is stranded, the honest answer takes a one-day assessment, not a hot take: workload inventory, voice dependency map, residency constraints, and a dated watch-list of Teams parity milestones specific to your estate. We run exactly that engagement, we operate both platforms daily, and we'll tell you plainly if the right move is to do nothing for a year. Get in touch.