Handle overwhelmed team as senior manager

# Step-by-Step Plan to Stabilize Execution and Well-Being ## 1) Diagnose Quickly and Objectively (48 hours) Goal: create a shared, data-driven picture of demand vs. capacity and where time is really going. - Inventory demand by type and source: - Buckets: roadmap/features, incidents/on-call, defects, tech debt, support/KTLO, compliance. - Pull from tracking tools (e.g., Jira): all in-flight and queued items; tag by type, owner, estimate, due date, dependency. - Build a simple capacity model: - Per engineer per week = 40h − meetings − on-call − interrupts − PTO. Example: 40 − 8 meetings − 4 on-call overhead − 4 interrupts = 24h effective. For 8 engineers → ~192h/week. - Compare to committed demand. If demand is ~280h/week, capacity gap is ~88h/week (46%). - Time and flow diagnostics: - Cycle time, throughput, WIP, queue age; percent unplanned work; context switching (tickets/engineer/week); on-call page rate, MTTR. - Calendar audit: meeting load, fragmentation (number of 30–60 min focus blocks). - Human signal check: - 1:1s and an anonymous 3-question pulse: “How sustainable is the current pace? What’s the top drain? What should we stop?” - Identify bottlenecks and single points of failure (SPOFs): code owners with long review queues, approval gates, environment constraints. Deliverable: one-page “State of the Team” with demand vs. capacity, key risks, and initial recommendations. ## 2) Prioritize and Sequence Work Use explicit tradeoff frameworks and limit work-in-progress to increase flow. - Choose a prioritization method: - WSJF (Weighted Shortest Job First): WSJF = Cost of Delay / Job Size. - Or MoSCoW: Must/Should/Could/Won’t for the current time window. - Establish a WIP limit (e.g., 1–2 active items per engineer) to reduce context switching. - Create a short, ordered backlog: - 0) Safety/regulatory/compliance and Sev-1/2 defects. - 1) Critical-path features for near-term external deadlines (MVP only). - 2) Risk-reduction/enablement work (infra, test, observability) that protects velocity. - 3) Nice-to-haves and deferred items. - Sequence by dependencies and critical path; use feature flags and phased rollouts to ship incrementally. Mini-example (WSJF): - A: Feature Alpha (CoD 100, Size 20) → 5.0 - B: Incident reduction (CoD 60, Size 6) → 10.0 - C: Dashboard nice-to-have (CoD 15, Size 3) → 5.0 Order: B, then A/C. If capacity only covers 26 “size” units this sprint, do B (6) + A (20) and drop C. ## 3) Redistribute Resources and Adjust Process - Skill matrix and load balancing: - Map engineers to skills; pair to de-risk SPOFs; move generalists to critical-path items; temporarily pause low-value streams. - Protect focus time: - Meeting cull; no-meeting blocks; async updates; batch code reviews; office hours for Q&A to reduce interrupts. - On-call and interrupts: - Cap on-call load per person; create a rotating “interrupt handler” to shield others; target <20% capacity for interrupts on average. - Short-term staffing levers: - Borrow from adjacent teams, short-term contractors, or reassign TPM/EM bandwidth for unblocking and coordination. ## 4) Negotiate Scope and Timelines Bring options, not problems. Translate capacity and flow data into business tradeoffs. - Option sets with explicit tradeoffs: - Scope cut (MVP): defer non-critical features; use flags; stage rollout. - Timeline shift: push dates with rationale (capacity gap, dependency latency). - Resource change: add headcount/loans or reduce parallel projects. - Communicate via a concise decision memo: - Current state, risks, 2–3 options with impact on outcomes, recommended path, and new milestones. - Freeze policy: - Temporary intake freeze on new requests unless they displace lower-priority work via the same framework. ## 5) Support Well-Being and Sustainable Pace - Culture and policy: - No-heroics norm; discourage after-hours except true incidents; rotate tough work fairly; recognize effort publicly. - Encourage PTO; provide mental health and EAP resources; set “quiet hours.” - Psychological safety: - Frequent 1:1s, skip-levels; celebrate small wins; blameless postmortems. - Guardrails: - Max WIP per person; overtime tracked; if overtime > 3–5h/week sustained, trigger scope/timeline review with stakeholders. ## 6) Communication Cadence and Artifacts - Kickoff reset (Day 2–3): live readout + doc covering priorities, WIP freeze, and what changes for whom. - Weekly stakeholder update (one page): - Plan vs. actual, top risks/asks, shipped items, next week’s focus, metrics snapshot. - Single source of truth: - Shared roadmap/backlog board with priorities; risk/decision log; status page with green/yellow/red per stream. - Incident and risk comms: - Clear SLAs, incident channel, and roles; pre-approved templates for external updates. ## 7) Measure Whether It’s Working (Leading and Lagging) Track a balanced set of flow, quality, delivery, and health metrics. - Leading indicators (weekly): - WIP per engineer (target ≤ 2), cycle time (e.g., median days to done), queue age (90th percentile), % unplanned work, focus time hours, context switches. - On-call: pages/week, MTTR, interrupt time %. - Team pulse: eNPS/likert on sustainability; overtime hours. - Lagging indicators (bi-weekly/monthly): - Throughput/velocity stability (within ±15%), on-time delivery to re-baselined plan, change failure rate, escaped defects. - Simple targets example (over 4–6 weeks): - Reduce median cycle time from 9d → 6d; cut unplanned work from 40% → 20%; keep WIP ≤ 2; maintain change failure rate < 20%; improve pulse “sustainable pace” from 2.8/5 → 4.0/5. - Review cadence: - Weekly metric review + monthly retro; if targets aren’t trending, revisit scope/timeline/resources first—not people’s hours. ## 8) Contingencies and Edge Cases - Hard external deadline (launch/regulatory): create a tiger team, freeze non-critical work, daily check-ins, and an explicit recovery plan post-deadline. - Dependency bottlenecks (e.g., another team): formalize SLAs, escalate via a shared steering forum, or decouple via mocks/adapters. - Quality risk: add test automation/observability as first-class items; block launches on minimum quality gates. ## 9) Example Timeline - Days 1–2: Data pull, 1:1s, capacity model, WIP freeze on non-critical work. - Day 3: Reset meeting; publish prioritized plan and metrics baseline. - Week 1–2: Implement WIP limits, interrupt handler, meeting cull, paired work on SPOFs. - Week 3–4: Re-baseline commitments with stakeholders; ship MVP slices; start phasing deferred scope. - Week 5–6: Review metrics and pulse; adjust; decide on ongoing staffing or scope changes. ## Common Pitfalls to Avoid - Treating burnout as an individual time-management issue instead of a system/capacity problem. - Adding people to a late project without decoupling work (Brooks’s Law). - Prioritizing by loudest stakeholder instead of a transparent framework. - Hiding risk; slipping dates silently; accumulating hidden WIP. This approach creates transparency, reduces WIP and interrupts, protects the team, and aligns delivery with realistic capacity while preserving stakeholder trust.