Posts

Retrospective formats compared: which one should your team use?

An agile team gathered around a whiteboard covered with colorful sticky notes organized in different retrospective formatsAn agile team gathered around a whiteboard covered with colorful sticky notes organized in different retrospective formats
Matt Lewandowski

Matt Lewandowski

Last updated 16/02/202613 min read

The format you choose for a retrospective matters as much as running the retro itself. A team that uses the same three-column board every sprint eventually stops seeing new insights. The prompts become wallpaper. People recycle the same feedback. The meeting feels mandatory rather than useful. The fix isn't to stop running retros. It's to match the format to the situation. Different formats surface different kinds of feedback. Some are better for new teams that need structure. Others work best when a team is stuck in a rut and needs a fresh lens. A few are designed specifically for teams that have trouble speaking up. This guide compares six of the most widely used retrospective formats, explains when each one shines, and gives you a decision framework so you can pick the right one without guessing.

The six formats at a glance

Before diving into the details, here's how the formats compare across the dimensions that matter most.
FormatColumns / CategoriesBest team sizeBest forLearning curve
Start-Stop-ContinueStart, Stop, Continue3-10New teams, quick retrosLow
4LsLiked, Learned, Lacked, Longed For4-12Reflective teams, milestone retrosLow
SailboatWind, Anchor, Rocks, Island4-15Visual thinkers, forward-looking teamsMedium
Mad-Sad-GladMad, Sad, Glad3-10Emotional check-ins, post-incidentLow
DAKIDrop, Add, Keep, Improve4-12Process-focused teams, mature ScrumLow
Lean CoffeeTo Discuss, Discussing, Discussed3-20Self-organizing teams, open agendasMedium


Start-Stop-Continue

How it works

The team answers three questions:
  • Start: What should we begin doing that we aren't doing now?
  • Stop: What should we stop doing because it's not helping?
  • Continue: What's working well and should keep going?
Each person writes sticky notes (physical or digital), the facilitator groups them, and the team votes on the most important items to discuss and act on.

When to use it

Start-Stop-Continue is the default retrospective format for a reason. It's simple, action-oriented, and requires zero explanation. Every sticky note maps directly to a behavior change, which makes it easy to turn discussion into commitments.

Strengths

  • Produces immediately actionable output
  • No learning curve for participants
  • Works well in short time slots (30 minutes)
  • Forces specificity because every item is framed as a behavior

Limitations

  • Doesn't capture emotions or learning, only actions
  • Can feel repetitive after many sprints
  • The "Continue" column often gets ignored because it feels like filler
  • Teams may struggle to distinguish "Start" from "Improve"


4Ls (Liked, Learned, Lacked, Longed For)

How it works

The team reflects across four dimensions:
  • Liked: What did you enjoy or appreciate during the sprint?
  • Learned: What new knowledge or insights did you gain?
  • Lacked: What was missing that you needed?
  • Longed For: What do you wish you had that doesn't exist yet?
Participants add items to each column, then the group discusses patterns and identifies actions.

When to use it

The 4Ls format is particularly strong at milestone moments: end of a release, end of a project phase, or after a team has been running for several sprints and wants to take a broader view. The "Learned" column is uniquely valuable because most other formats skip reflection entirely.

Strengths

  • Captures positive experiences alongside gaps
  • The "Learned" column surfaces growth that would otherwise go unacknowledged
  • "Longed For" encourages aspirational thinking rather than just problem-fixing
  • Naturally balanced between backward-looking and forward-looking

Limitations

  • Less immediately action-oriented than Start-Stop-Continue
  • "Lacked" and "Longed For" can overlap, confusing participants
  • Requires more facilitation to convert reflections into concrete action items
  • Some teams find it too introspective for a regular sprint retro


Sailboat

How it works

The team uses a sailboat metaphor to structure their feedback:
  • Wind (sails): What's propelling the team forward?
  • Anchor: What's slowing the team down or holding it back?
  • Rocks (risks): What dangers or obstacles lie ahead?
  • Island (goal): What is the team working toward?
The facilitator draws a sailboat on the board (or uses a digital template) and team members place sticky notes in each area. The visual nature of the format tends to generate more creative, less formulaic responses.

When to use it

The Sailboat format stands out for teams that are stuck producing the same feedback sprint after sprint. The metaphor gives people permission to think differently about familiar problems. The "Rocks" category is especially useful because it's explicitly forward-looking, which most retrospective formats aren't.

Strengths

  • Visual format breaks the monotony of column-based retros
  • The "Rocks" category surfaces risks before they become problems
  • The "Island" keeps the team aligned on shared goals
  • Works well with larger teams because the visual is engaging

Limitations

  • Takes more setup time (drawing the boat, explaining the metaphor)
  • Some participants find metaphorical categories harder to map to real issues
  • Less structured than column formats, so facilitation matters more
  • Can feel forced if the team isn't receptive to visual exercises


Mad-Sad-Glad

How it works

The team sorts their feedback by emotional response:
  • Mad: What frustrated or angered you?
  • Sad: What disappointed you or made you feel let down?
  • Glad: What made you happy or proud?
This format explicitly centers emotions rather than actions or processes. The facilitator needs to create psychological safety for this to work, since people are being asked to be vulnerable about how they feel, not just what they observed.

When to use it

Mad-Sad-Glad is the right choice when the team has been through something emotionally significant: a failed release, a period of high stress, a conflict, or a major win. It gives people a structured way to express what they're feeling before jumping to problem-solving.

Strengths

  • Surfaces the human side of teamwork that process-focused formats miss
  • Validates emotions, which builds trust and team cohesion over time
  • Simple enough that no one needs instructions
  • The "Glad" column reinforces positive behaviors and celebrates wins

Limitations

  • Requires high psychological safety to be effective
  • Can feel uncomfortable for teams that aren't used to discussing emotions at work
  • Doesn't naturally produce action items without strong facilitation
  • Not ideal as the only format used repeatedly, since it can become emotionally draining


DAKI (Drop, Add, Keep, Improve)

How it works

The team evaluates their current practices through four lenses:
  • Drop: What should we stop doing entirely?
  • Add: What new practice should we introduce?
  • Keep: What's working well and should stay?
  • Improve: What exists but needs to be better?
DAKI is similar to Start-Stop-Continue but adds a critical distinction: the difference between something that needs to be added from scratch versus something that exists but needs improvement. This nuance helps mature teams fine-tune their processes rather than making blunt start/stop decisions.

When to use it

DAKI works best for teams that have been working together long enough to have established practices worth evaluating. It's particularly effective when a team is going through a process change or wants to audit their current way of working.

Strengths

  • More granular than Start-Stop-Continue
  • The "Improve" category acknowledges that some things are partially working
  • Every item maps to a clear action type
  • Works well for process audits and quarterly reviews

Limitations

  • Very similar to Start-Stop-Continue, so teams may not feel a significant difference
  • Can be overly process-focused at the expense of team dynamics
  • "Drop" and "Improve" can overlap when something is mostly broken
  • Less useful for new teams that haven't built enough process to evaluate


Lean Coffee

How it works

Unlike the other formats, Lean Coffee doesn't have predefined categories. Instead:
  1. Build the agenda: Everyone writes topics they want to discuss on sticky notes
  2. Vote: The team dot-votes on which topics matter most
  3. Discuss: Topics are discussed in priority order, with a timer (usually 5 minutes per topic)
  4. Extend or move on: After the timer, the team votes thumbs up/down to extend discussion or move to the next topic
The board has three columns: To Discuss, Discussing, and Discussed.

When to use it

Lean Coffee is the best choice when the facilitator isn't sure what the team needs to talk about, or when the team is self-organizing enough to drive their own agenda. It respects that the team knows what matters most.

Strengths

  • The team decides what matters, not the format
  • Democratic: voting ensures the most pressing topics get airtime
  • Timeboxing prevents any single topic from dominating
  • Works well for large groups or cross-team retrospectives

Limitations

  • No built-in structure to ensure balanced feedback
  • Requires a self-organizing team; quiet members may not propose topics
  • Can feel unstructured for teams that prefer clear prompts
  • Doesn't naturally produce categorized action items


Decision framework: pick the right format for the situation

Rather than defaulting to the same format every sprint, match the format to what's actually happening on your team.
SituationRecommended formatWhy
New team, first retrosStart-Stop-ContinueZero learning curve, immediately actionable
End of release or project phase4LsCaptures learning and aspiration alongside problems
Team is stuck giving the same feedbackSailboatMetaphor breaks habitual thinking patterns
After a stressful sprint or incidentMad-Sad-GladNames emotions before jumping to fixes
Mature team refining existing processesDAKIDistinguishes between adding new and improving existing
Team wants to drive their own agendaLean CoffeePuts the team in control of what gets discussed
Large group (15+) or cross-teamLean Coffee or SailboatBoth scale well; Lean Coffee via voting, Sailboat via visual engagement
Team with low psychological safetyMad-Sad-Glad (with anonymous input)Emotional framing makes it safer to surface concerns
If you're unsure where to start, try generating a format with Kollabe's AI-powered retrospective template generator. It builds a customized board for your specific context.

How to keep retrospective formats fresh

Even the best format gets stale after enough repetitions. Here are proven strategies for keeping your retros engaging.

Rotate formats intentionally

Don't rotate randomly. Instead, choose the format based on what happened in the sprint. A difficult sprint calls for Mad-Sad-Glad. A milestone calls for 4Ls. A routine sprint can use Start-Stop-Continue. The rotation happens naturally when you match format to context rather than following a calendar.

Let the team choose

At the end of each retro, ask: "What format should we use next time?" Give them three options and let them vote. This small act of ownership makes the next retro feel more relevant before it even starts. Browse Kollabe's template library to show the team what's available.

Start with a warm-up

The first five minutes of a retro set the tone for everything that follows. A quick icebreaker question can shift people from "meeting mode" to "reflection mode." Kollabe's icebreaker generator provides hundreds of options you can use without any prep.

Adapt the categories

No format is sacred. If the Sailboat's "Rocks" category doesn't resonate, rename it to "Hazards ahead" or "Watch out for." If your team consistently ignores the "Continue" column in Start-Stop-Continue, drop it and add "Celebrate" instead. The categories exist to prompt thinking, not to enforce a rigid structure.

Combine formats

Some teams blend elements from multiple formats. For example, opening with Mad-Sad-Glad to surface emotions, then switching to DAKI to plan specific improvements. This takes more facilitation skill but can be highly effective for teams that have outgrown any single format. For more creative approaches beyond these standard formats, check out our guide on agile retrospective ideas that covers themed and unconventional retro activities.

Common mistakes when choosing a format

Using a complex format with a brand-new team. Sailboat and Lean Coffee require facilitation confidence and team maturity. Start with Start-Stop-Continue or Mad-Sad-Glad and add complexity as the team builds trust. Switching formats every sprint without reason. Rotation for its own sake prevents the team from building fluency with any single format. Stick with a format for 3-4 sprints before moving to a new one, unless the situation clearly calls for something different. Choosing a format that avoids the real issue. If the team has an emotional problem (burnout, conflict, frustration), don't use a process-focused format like DAKI. Use Mad-Sad-Glad to address what people are feeling first. Skipping the action items. Every format needs to end with clear commitments. The format is just the input mechanism. Without a shared agreement on "what are we actually going to do differently," even the best retro is just a venting session. If you want a complete guide to running an effective agile retrospective from start to finish, including facilitation tips that apply to any format, we've covered that separately.

Frequently asked questions

Change the format when it stops producing new insights, not on a fixed schedule. Most teams find that rotating every 3-5 sprints keeps things fresh without losing the benefit of familiarity. The exception is when a significant event (incident, release, team change) makes a specific format clearly more appropriate.

Yes, and experienced facilitators often do. A common approach is to use Mad-Sad-Glad for the initial data gathering phase to surface emotions, then switch to DAKI or Start-Stop-Continue for the action planning phase. Just be careful not to over-engineer the session. Two formats combined should still feel simpler than one format done poorly.

All six formats work remotely with the right tooling. Start-Stop-Continue and DAKI translate most naturally to digital whiteboards because they're column-based. Sailboat can work well if your tool supports drawing or visual templates. Lean Coffee is excellent for remote because the voting mechanism keeps large groups focused. The key for remote retros is anonymous input, which helps compensate for the reduced social pressure to participate.

Resistance to new formats usually signals one of two things: the current format is working well (in which case, don't fix what isn't broken), or the team doesn't feel safe enough to try something unfamiliar. If it's the latter, introduce the new format as a one-time experiment rather than a permanent change. Frame it as "let's try this for one sprint and see what happens." Most teams are willing to experiment when the commitment is low.