Teams Phone Product Management - Balancing Innovation and Fundamentals

Prefer audio?

Who should listen: Microsoft 365 service owners and UC professionals running or planning Teams Phone deployments who want to understand how calling features are prioritised inside Microsoft, why parity gaps persist, and where AI is genuinely heading for calling rather than meetings.

Guest: Emily Kirby George — Senior Product Manager, Teams Calling, Microsoft

Emily Kirby George is a Senior Product Manager on the Microsoft Teams calling team who came into product via eight years in Skype for Business product support and Microsoft IT's internal Teams migration. She explains how Teams Phone features get prioritised across three pillars — fundamentals, AI and long-tail parity — why hard cuts happen, and how AI for calling is being designed differently from meetings.

Many thanks to Logitech, sponsor of this episode.

Key insights

• Teams Phone is now past 20 million monthly active users, but Arbuthnot notes the far larger remaining base sitting on legacy PBXs and other platforms is what the roadmap has to win over. ▶ 7:52
• Delayed Simultaneous Ring — the ability to hold ringing of the rest of a call group for around 15 seconds — is a good example of an unglamorous long-tail feature with no AI in it that large customers quantified in lost productivity, and it directly unlocked several migrations. ▶ 8:53
• Small features rarely reach all users, but they remove the specific blocker stopping an entire organisation moving; Kirby George confirmed several customers were unblocked by Delayed Simul Ring alone. ▶ 10:28
• Calling investment is prioritised in three buckets in order: fundamentals and table stakes first, then AI/intelligence to modernise the business, then the long tail to round out the feature set. ▶ 12:30
• Semester planning uses explicit 'hard cuts' — a feature ships or it is cut — and when the team visualised planned hours for the first half of 2025, the fundamentals bucket was by far the fullest, ahead of the AI work. ▶ 12:00
• AI for calling is deliberately not a copy-paste of meetings Copilot: calls have no calendar entry and no pre/during/post lifecycle or artefacts, so recap-style features (e.g. 'catch me up on the last five minutes') don't map across. Focus areas instead include queues app intelligence for agent and queue performance, B2B scenarios and pulling caller/CRM context into the call. ▶ 14:36
• The reliability work behind the scenes includes the service moving to a five-nines commitment and Survivable Branch Appliance improvements such as the seven-day token — the majority of the team's effort goes into these non-headline investments. ▶ 17:41
• Cross-team dependencies are a real constraint, not an excuse: one of Kirby George's current features needs seven or eight separate engineering teams with their own priorities to align, which is why obvious-looking UserVoice items stay open for years. ▶ 20:44

Insights summarised by AI from the episode transcript, reviewed by the Empowering.Cloud team.

Listen: Apple Podcasts · Spotify · Other platforms