Identify Systemic Issues From 3 Months of Retro Notes
Analyzes 3 months of retrospective notes to surface systemic issues that recur across sprints. Built for Scrum Masters doing quarterly team health reviews.
When to use this prompt
Use this at the end of a quarter when you want to see whether your retrospectives are actually changing anything. You will need 8-12 sprints worth of retro notes (stickies, action items, summaries) and the list of action items committed in each retro. The prompt looks for patterns of issues that keep appearing despite action items, which is the clearest signal of a systemic problem that sprint-level retros cannot fix. It is most valuable for mature teams that have been running retros consistently for at least a quarter. New teams will not have enough data yet.
The Prompt
You are a Scrum Master and team coach analyzing 3 months of retrospective notes to identify systemic issues. Your job is to distinguish sprint-level hiccups from deeper patterns that recur across multiple sprints. Retrospective notes (last 3 months): {{retro_notes}} Action items committed in each retro: {{action_items}} Quarter name: {{quarter}} Team: {{team_name}} Produce the analysis in this structure: 1. FREQUENCY MAP â For each theme mentioned across retros, count how many sprints it appeared in. Order by frequency. Themes appearing in 5+ sprints are candidates for systemic issues. 2. ACTION ITEM TRACE â For each systemic theme, trace the action items that were committed to address it. For each action item, note: - Was it completed? - Did it change the underlying issue? - Did the theme reappear after the action item was committed? 3. SYSTEMIC DIAGNOSIS â For themes that appeared 5+ times AND had action items that did not resolve them, run a 3-layer root cause analysis: - Symptom: what the team names in retro - Immediate cause: the obvious proximate reason - Systemic cause: the deeper structural or organizational reason 4. BEYOND TEAM CONTROL â Flag systemic causes that the team cannot fix on its own. These need escalation to engineering leadership, HR, or executive level. 5. RECOMMENDED QUARTER ACTIONS â 2-3 specific actions for the next quarter, each targeting a systemic issue rather than a symptom. For each: - What will change - Who owns it (including the external stakeholder if needed) - How we will know it is working Rules: - Honest about what is not fixable at the team level. Teams waste energy trying to solve leadership-level problems from the bottom up. - If the same action item has been committed and re-committed across 3+ sprints without change, explicitly say so; that is diagnostic information. - If the team has no systemic issues and all retros produced genuine improvement, say that honestly. Manufactured findings are worse than "things are going well."
Example Output
FREQUENCY MAP (last 10 sprints) - PR review bottleneck: 9 sprints - Standup over-runs: 7 sprints - Unclear definition of done: 6 sprints - Tech lead burnout: 6 sprints - Sprint scope creep: 5 sprints - On-call fairness: 4 sprints ACTION ITEM TRACE - PR review bottleneck: 4 action items committed, 3 completed (add second reviewer, reduce review checklist, pair reviews on Fridays), 1 in flight. Theme continued appearing despite completion. - Standup over-runs: 2 action items (timebox, enforce 2-min cap), both completed. Theme still appears but less intensely. SYSTEMIC DIAGNOSIS - PR review bottleneck: Symptom: reviews take 2-3 days. Immediate cause: tech lead is the only approved reviewer on critical paths. Systemic cause: team has not invested in knowledge distribution; tech lead is a single point of failure by organizational design, not choice. - Tech lead burnout: Symptom: tech lead working late, missing personal time. Immediate cause: tech lead is overloaded with reviews, architecture questions, and incident response. Systemic cause: we have grown 2x in 6 months without adding a second senior engineer. Staffing ratio is wrong. BEYOND TEAM CONTROL - Staffing ratio (tech lead burnout) is a hiring decision owned by engineering leadership. - On-call fairness needs people ops involvement. RECOMMENDED QUARTER ACTIONS 1. Formalize knowledge sharing sessions every Friday so PR review knowledge is distributed, not just dependent on a single person. Owner: PM + Tech Lead. Success: at least 3 engineers can approve billing service PRs by end of quarter. 2. Escalate tech lead staffing to engineering leadership with frequency data. Owner: PM. Success: commitment to hire or reassign by end of next month. 3. Propose quarterly on-call rotation review with people ops. Owner: PM + People Ops. Success: rotation policy reviewed by end of quarter.
Recommended Tools
Notion projects is where most teams keep retro notes, so the analysis can be run directly against the source. Dovetail is purpose-built for thematic analysis of text data and preserves the evidence chain for each theme. Airtable works when retros are stored in a structured database. Use Notion if your retros are documents; Dovetail or Airtable if they are in a more structured format.
Frequently Asked Questions
When should I use this prompt?
Use it quarterly, at the transition between quarters, as part of a team health review. It is especially valuable when the team has been running retros consistently for at least 8-12 sprints and there is a suspicion that the same issues keep recurring. Do not use it monthly; the signal-to-noise ratio is too low. Also do not use it on teams new to retrospectives (fewer than 6 months of history) because you do not have enough pattern data to distinguish systemic from situational. Quarterly cadence is the sweet spot for this analysis.
What if the analysis reveals a leadership problem?
That is often exactly what happens, and it is the most valuable outcome of running this prompt. Teams consistently rediscover that their recurring issues are caused by staffing, priorities, or organizational structure decisions above their level. When this happens, the right response is to escalate with data: show the frequency map, the action item trace, and the systemic diagnosis to engineering leadership in a structured conversation. Leaders respond better to evidence-based escalations than to complaints, and the analysis gives you the evidence. Do not use the findings to vent; use them to propose specific changes leadership can approve.