← Agile/Scrum Suite 23 · Governance & Stakeholder

Program Closeout & Lessons Learned

Download Word
✅ Program Closed — May 15, 2026 · Release 2 live, CareLink Classic fully decommissioned

Final Performance Summary

DurationSprint 0 + 8 sprints, 18 wks
Scope delivered337 / 342 pts (98.5%)
Budget$1.81M / $1.85M (97.8%)
Avg. velocity43 pts/sprint
Predictability~94%
Critical defects at launch0

Deliverable Acceptance

DeliverableAcceptance
Release 1 (MVP) — patient mobile app, provider console coreAccepted, Mar 27, 2026 (M. Delacroix)
Release 2 — messaging, refills, full migration, cutoverAccepted, May 8, 2026 (M. Delacroix)
Full historical patient data migrationValidated & accepted (T. Brannigan)
CareLink Classic decommissionComplete, verified May 12, 2026

Outstanding Items

Lessons Learned

1. Linked ≠ Watched — Cross-Team Dependencies Need an Owner

Tracking a dependency in the tool isn't the same as someone actively watching it. The Sprint 2 near-miss on MCM-111 led to adding an explicit dependency-owner practice, which held for the rest of the program.

Traced to: Sprint 2 Retrospective, RAIDD Log R-03

2. Splitting Oversized Stories Early Pays Off

MCM-133's split into a data-mapping story and a UI story (Sprint 3 refinement) avoided what likely would have been a Sprint 6 delivery risk — the pattern held up well enough that both teams applied it proactively on later large stories without prompting.

Traced to: Sprint 3 Backlog Refinement, RAIDD Log DEC-03

3. Staged Compliance Checkpoints Beat a Single End-of-Program Gate

Splitting the compliance review into a Release 1 checkpoint and a Release 2 checkpoint (decided at Kickoff) meant migration validation issues, had any occurred, would have surfaced with runway to fix them — rather than all validation risk landing in the final sprint.

Traced to: Program Kickoff, RAIDD Log DEC-02

4. Phased Rollout After MVP Reduced Support Risk

Holding Release 1 to the pilot cohort rather than immediate full rollout (Sponsor decision at Sprint 5 Review) kept early support ticket volume manageable and gave the team a smaller blast radius for any post-launch issues — none of which materialized, but the decision was sound regardless of that outcome.

Traced to: Sprint 5 Review, RAIDD Log DEC-04

5. A Two-Team Scrum-of-Scrums Model Scaled Well at This Size

Two teams of five was small enough that a lightweight, twice-weekly Scrum-of-Scrums was sufficient coordination overhead — a heavier SAFe-style Agile Release Train structure would likely have been more process than this program's size warranted.

Traced to: Team Charter, program retrospective input from both Scrum Masters

Recommendation for Future Programs

This delivery model (two coordinating Scrum teams, lightweight Scrum-of-Scrums, staged compliance checkpoints) is recommended as a repeatable pattern for future ACME Health digital programs of similar size — before reaching for a heavier scaled-Agile framework.