Start Building

Lesson 46: Quarterly Recurrence Audit Pack for Release Windows and Capacity Planning in RPG Live-Ops

Lesson 45 gave you class-level weekly trend rollups. The next step is turning those weekly signals into a quarterly decision pack that informs release windows and staffing lanes before risk accumulates silently.

This lesson builds a compact quarterly audit workflow your release owners can use during planning and go or hold discussions.

Cozy House illustration for quarterly recurrence audit lesson

What you will build

By the end of this lesson, you will have:

  1. A quarterly_recurrence_audit_pack.md structure with fixed sections and evidence links
  2. A quarterly_recurrence_summary.csv table mapping risk classes to release windows
  3. A capacity pressure matrix that ties unresolved recurrence to owner bandwidth
  4. A release-window recommendation rubric (proceed, proceed_with_watch, hold)

Step 1 - Build the quarterly pack skeleton

Create these required sections:

  • Quarter scope and release window boundaries
  • Recurrence class trend summary (import from Lesson 45 data)
  • Open escalation items and age
  • Capacity and owner coverage by lane
  • Final release-window recommendation and sign-off

Keep section order stable every quarter so leadership compares like-for-like packets.

Step 2 - Author quarterly_recurrence_summary.csv

Minimum columns:

column purpose
quarter_id e.g. 2026-Q3
recurrence_class normalized class key from weekly rollup
avg_weighted_risk_score quarter average of weekly risk
peak_week_score highest observed weekly score
open_reopen_count unresolved reopened incidents in class
overdue_action_count unresolved items past SLA
owner_lane accountable lane
capacity_state healthy, stretched, critical
window_recommendation proceed, proceed_with_watch, hold

Pro tip: lock formula fields in one shared template to prevent quarter-to-quarter score drift from spreadsheet edits.

Step 3 - Add capacity pressure scoring

For each owner lane, score pressure with three signals:

  1. unresolved recurrence actions
  2. release-critical tasks in the same window
  3. backup-owner coverage depth

If any lane is critical, your default recommendation should not be automatic proceed, even if class trend scores are improving.

Step 4 - Tie recommendations to release windows

Use one simple rule set:

  • proceed when all classes are green or stable yellow with healthy capacity
  • proceed_with_watch when yellow classes have bounded action plans and explicit due dates
  • hold when red class risk or critical capacity pressure overlaps the same release window

Always link each recommendation to concrete lineage references (watch_id, parent_incident_id, closure_packet_version) so decisions are auditable.

Step 5 - Publish the audit packet to decision lanes

Share one packet with:

  • release captain
  • QA lead
  • support or community lane owner
  • legal or compliance lane when applicable

Attach one short summary paragraph per lane so no team has to decode the full packet during planning meetings.

Common mistakes

Mistake: Treating quarterly audit as a narrative memo

Fix: keep it structured. Tables first, commentary second.

Mistake: Ignoring capacity state because trend is improving

Fix: capacity is a risk multiplier. Improving trend with critical staffing is still fragile.

Mistake: Aggregating too many classes in one recommendation row

Fix: keep recommendations class-aware and then produce one portfolio summary row.

Pro tips

  • Keep one unchanged recommendation rubric across the year
  • Track audit packet publish timestamp and approval timestamp separately
  • Require explicit owner acknowledgements for all hold and proceed_with_watch calls

Mini challenge

  1. Draft one quarterly_recurrence_summary.csv row for two classes in the same quarter.
  2. Mark one lane stretched and one lane critical.
  3. Write a one-line release-window recommendation with linked lineage ids.

FAQ

Does this replace weekly recurrence review

No. Weekly review drives operational reaction. Quarterly audit drives release planning and capacity policy.

How many recommendation states should we keep

Three states are enough for most small teams: proceed, proceed_with_watch, and hold.

What if leadership asks for a single score only

Provide a single portfolio row as a summary, but keep class-level and capacity data visible in the packet.

Lesson recap

You now have a quarterly recurrence audit pack that turns weekly risk signals into release-window and staffing decisions with clear accountability and lineage traceability.

Next lesson teaser

Continue with Lesson 47: Cross-Quarter Variance Review for Release Window and Staffing Capacity in RPG Live-Ops to compare recurrence trends across quarters and tie recommendation shifts to staffing responses.

Related learning