Technical Program Manager – Program Management

Zero-to-one program execution (Platform Delivery): How would you launch a complex cross-functional program from the ground up—for example, a new internal platform—across 100+ engineering teams in multiple business units without disrupting core product delivery?

At 100+ teams across multiple business units, a program can no longer rely on a single executive or a per-team champion network — the coordination structure itself has to scale sub-linearly with team count. The shift is from mandating adoption into every backlog to building a platform good enough, and self-service enough, that teams choose to adopt it faster than they could be forced to.

  1. 1.Secure sponsorship at the BU level, not just the top — get a named VP/Director sponsor inside every business unit, fund a dedicated central platform team upfront, and tailor the business case to what each BU actually values.
  2. 2.Segment the 100+ teams instead of treating them as one population — inventory every team by BU, stack, and migration complexity, and co-design the platform with 2–3 lighthouse teams against a real production workload.
  3. 3.Build a two-tier governance structure — a central platform team owns the roadmap and interface contracts, one platform lead per BU coordinates adoption locally, and a lightweight architecture review board approves cross-BU-impacting changes.
  4. 4.Roll out in adoption waves, gated by exit criteria — prove the platform with lighthouse teams in Wave 0, expand to one full BU in Wave 1 once exit criteria are met, then roll to remaining BUs in parallel, paced by the central team's own capacity.
  5. 5.Make adoption self-service and pull-based — invest disproportionately in a self-service developer portal and automated migration tooling, make the new path strictly better on a metric teams already care about, and support dual-run with clean rollback.
  6. 6.Run a scaled support and visibility cadence — staff standing office hours and a shared support channel, track adoption and migration velocity per BU on a visible dashboard, and run a biweekly cross-BU sync to unblock dependencies.

This holds up at 100+ teams because both the coordination structure (BU leads, not per-team champions) and the adoption mechanism (self-service pull, not mandated push) scale sub-linearly with team count — freeing the central team's limited bandwidth for the handful of BUs that genuinely need help instead of chasing every team individually.