methodatlas
RunsheetAgile

Sprint Retrospective

ComplexityLow
Time45-90 min
Participants3-10
FormatWorkshop
MaturityCanonical
01

Prerequisite

What needs to be finished first

Complete firstSprint completionnot in catalog

The current sprint is completed with review and increment demo, so concrete experiences exist.

Without: Without completed sprint, the team discusses speculation instead of lived experience and actions become vague.
Complete firstBlameless Postmortem

Psychological safety is established or explicitly requested in the session, so critical points can be voiced.

Without: Without this norm, the retro stays polite and nothing improves because the real topics are not addressed.
02

Preparation

What needs to be ready before start

Materials

Miro or FigJam board with template (for example Mad-Sad-Glad, Sailboat, Start-Stop-Continue); timer; action board for measures with owner and deadline; list of action items from previous retro.

People / roles

One facilitator (rotating or Scrum Master); all development team members; optionally Product Owner; no external stakeholder without invitation.

Pre-read

Sprint burndown, velocity, completed and uncompleted items; major incidents in sprint; status of previous action items; shared initial mood (pulse check).

Time needed

60-90 min per 2-week sprint

Setup

Vary format per retro to avoid routine. Make previous action items visible. State rules: Vegas rule (what is said here stays here), focus on system instead of people, one action beats ten.

03

Core question

The one question this method answers

Which one change to process or system will improve the next sprint most, and who will implement it bindingly?

04

Flow

Marker: Phase

StepDurationActionHint
1Phase 1: Set the Stage
5-10 minName purpose of retro, repeat rules, pulse check (thumbs or 1-5). Briefly review previous action items: done, open, dropped.If the same action items are unresolved across three retros, that is the only meaningful topic for this retro.
2Phase 2: Gather Data
15-20 minFormat-specific collection: Mad-Sad-Glad, what went well/badly, 4Ls, Sailboat. Silent writing first, then cluster. Include metrics (velocity, bug count, incident count).If everyone repeats the same phrase in a format, topic breadth is too narrow. Change format or reframe question, otherwise retro becomes obligation exercise.
3Phase 3: Generate Insights
15-20 minVote top topics (dot voting, 2-3 points per person). Use 5 Whys or Fishbone on top topic to find cause. Work deeply on maximum two topics.One topic with depth beats five superficial ones. If everything is important, nothing changes.
4Phase 4: Decide What to Do
15-20 minOne action per top topic: SMART wording, with owner, deadline and success indicator. Maximum two to three action items per retro.Action item without owner is a wish. Owner must be present and say yes, otherwise it does not go on the list. Success indicator must be checkable in next retro.
5Phase 5: Close the Retro
5 minPlus-Delta or round robin: what was helpful in this retro, what should differ next time. Summarize action items, confirm responsibility.Closing is not "bye". If retro value is unclear, next engagement drops. Plus-Delta provides improvement loop for the retro itself.
05

Artifact

What comes out at the end

Form

Retro documentation with sprint ID, participants, collected topics (clusters and bullets), top insights with root-cause analysis, action items (description, owner, deadline, success indicator), Plus-Delta notes.

Versioning / ownership

One entry per sprint with date and sprint ID. Track action items with unique ID, status (open, in progress, done, dropped) across retros. Quarterly review of action-item completion rate as meta indicator.

Tool alternatives
  • Miro or FigJam with retro templates
  • Retrium or EasyRetro for remote workshops
  • Confluence or Notion template with embedded action board
  • Parabol with integrated voting and action tracker

sprint-retrospective-working-template.md

Compact working template for Sprint Retrospective with context, input, output artifacts, and next step.

Sprint Retrospective Working Template

Goal

A Scrum event to inspect and adapt collaboration, quality, and process.

Context

When and for what do we use this method?

Input

Which data, observations, decisions, or materials are available?

Execution

Short notes along the runsheet.

Output artifacts

  • Improvement Actions:
  • Team Agreements:

Assumptions and open questions

  • ...

Decision / Next step

Owner, date, and success signal.

06

Example output

Concrete filled scenario, fictional example

sprint-retrospective-beispiel.md

Concrete filled scenario, fictional example

Sprint Retrospective - Platform Team Workspot, Sprint 47 (CW 17-18)

Participants: 6 engineers, 1 PO, 1 Scrum Master.

Previous action items

  • AI-46-01: Extend Definition of Done with performance check. Status: done (PR #2841).
  • AI-46-02: Code review SLA 24 h. Status: open, brought into this retro.

Gather (Sailboat format)

  • Wind: Pair-programming sessions Tue/Thu closed knowledge gaps.
  • Anchor: Code reviews take >48 h, items stuck in review.
  • Rocks: External audit team blocks auth migration.
  • Island: 95th-percentile lead time under 10 days.

Top insight: Reviewer capacity is bottleneck. 5 Whys: current sprint load is 9.5 items per person, reviewers have no blocked time.

Action Items

  • AI-47-01: Daily review slot 14:00-15:00, owner @lisa, from Sprint 48. Indicator: 85th-percentile review wait time under 12 h.
  • AI-47-02: Lower WIP limit "In Review" to 3, owner @ben, from Sprint 48. Indicator: no card backlog in 3 consecutive Daily Stand-ups.

Plus-Delta: Plus: concrete action with indicator. Delta: review previous action items earlier.

07

Pitfalls

Recognize symptoms and steer against them

Trap

Repeated formats create routine

Symptom

Team arrives with same keywords, energy is low.

What to do

Vary format (Sailboat, 4Ls, Mad-Sad-Glad, Lean Coffee). Occasionally bring external facilitator eyes. Stretch timebox instead of shortening.

Trap

Too many action items

Symptom

Seven measures are agreed, by next sprint two are done, five lie fallow.

What to do

Maximum three action items, preferably one with large effect. Reject items without present owner. Watch completion rate as meta indicator.

Trap

Blame instead of system

Symptom

"X was too slow" or "Y did not see the bug" appears on cards.

What to do

Facilitator reframes: what in the system allowed the error. Person statements become process or tool statements. Otherwise psychological safety tips.

Trap

Action items without indicator

Symptom

"We improve reviews" without measurable definition. In three sprints unclear whether it worked.

What to do

Indicator mandatory. What would be evidence of success in 1-2 sprints. If not definable, cut action item differently.

Trap

Stakeholder in room

Symptom

Manager or customer sits in, team stays silent on sensitive topics.

What to do

Retro is team event. Externals only with explicit invitation and topic relevance. If escalation is needed, separate session afterward.

Trap

Previous action items ignored

Symptom

Items from three retros are open, nobody addresses it.

What to do

First agenda item: status of previous items. If open for three sprints, the only retro topic is "why do we close no action items".

08

Stop criteria

Done signals checkable in under a minute

Sprint was canceled without Increment, data basis missing.
Team does not trust confidentiality, statements are carried outward.
Manager insists on attendance against team's will, safety at risk.
Previous action items have not been processed for three retros, new items would land in same drawer.
Sprint data (velocity, bugs, incidents) unavailable, discussion stays anecdotal.
Team has newly formed, shared experience is missing.

Finished the runsheet?

Go to the profile for purpose, similar methods, and sources or continue to the next method in the catalog.