Starfish Retrospective Template: Start, Stop, More, Less, Keep

Introduction

A good retrospective template gives the team enough structure to think clearly without forcing a predetermined answer. Teams often discuss a practice as simply good or bad when the real issue is dosage.

This guide explains the approach in practical terms: what it is, why it is useful, when to use it, how to apply it step by step, and which mistakes to avoid.

What is it?

The Starfish retrospective uses five categories: Start Doing, Stop Doing, More Of, Less Of, and Keep Doing. It expands Start Stop Continue by distinguishing between practices that should change in intensity and practices that should be introduced or removed completely.

Why is it useful?

Teams often discuss a practice as simply good or bad when the real issue is dosage. More refinement may help, while too much refinement may create waste. The Starfish format supports those nuanced adjustments.

Benefits

• Distinguishes complete change from changes in frequency or intensity.
• Helps optimize existing practices rather than replacing them prematurely.
• Creates a balanced view of strengths and problems.
• Supports process tuning for mature teams.

When should you use it?

Use Starfish when the team has stable practices but needs to adjust them, when Start Stop Continue feels too binary, or after a process change that should be calibrated. It can become crowded for very large groups, so use strict writing and voting timeboxes.

How to do it

1. Explain the five choices. Give a simple example so participants understand the difference between Start and More, or Stop and Less.
2. Collect cards silently. Ask for specific behaviors, meetings, tools, and working agreements.
3. Clarify intensity. For More and Less cards, ask how much change would be useful and what limit should apply.
4. Group related practices. A single practice may appear across several categories. Use that contrast to clarify what the team actually wants.
5. Vote on impact. Prioritize the few adjustments most likely to improve the next cycle.
6. Define the experiment. Specify the new frequency, limit, owner, and review date.

Real example

A team placed code review in Keep, More, and Less. Discussion revealed that members wanted more early feedback on architecture but fewer large final reviews. The action was not simply do more reviews. The team introduced optional design comments before implementation and limited pull requests to a smaller size.

Common mistakes

• Treating the five columns as independent lists instead of comparing them.
• Using vague nouns without describing the behavior to change.
• Failing to quantify More or Less.
• Selecting several adjustments that compete with one another.

FAQ

Is Starfish suitable for new teams?

It can be, but Start Stop Continue may be easier for a team's first retrospective.

What if the same item receives votes in opposite categories?

That is useful information. Discuss the context, timing, or intensity that makes the practice helpful in one situation and harmful in another.

How many actions should result?

Usually one or two. The five columns expand insight, not the size of the action list.

Conclusion

Starfish is effective when improvement requires tuning rather than wholesale change. Use the five choices to understand intensity, then define a precise experiment the team can evaluate.

Call To Action

Create a Starfish template in SprintsPlans, group competing views of the same practice, and use voting to select the adjustment with the highest value.