KALM
Keep, add, less, more—fine-grained tuning of team habits.
Keep
What should remain as-is because it works.
Add
What is missing and should be introduced?
Less
What should happen less often or with lighter touch?
More
What should happen more often or with greater investment?
This board is for demo purposes only. Your responses are not saved. Close or refresh the page to clear all cards. Do not add any sensitive information.
What is KALM?
KALM stands for Keep, Add, Less, and More—a four-category retrospective format that focuses on adjusting the intensity and composition of team practices. The format emerged as a streamlined alternative to the five-category Starfish retrospective, removing the "Stop" column in favor of a more nuanced approach. KALM assumes that most practices have some value and focuses on calibration rather than elimination.
The four categories represent two types of decisions. "Keep" and "Add" are about what the team does: preserving existing practices and introducing new ones. "Less" and "More" are about how much the team does: adjusting the frequency, intensity, or investment in current activities. This dual-axis framing—what versus how much—gives teams a precise vocabulary for process optimization.
KALM is particularly effective because it avoids the harsh finality of "Stop." In practice, few team practices are entirely worthless. Code reviews that take too long are not worthless—they need to be streamlined, done less frequently, or done differently. By focusing on calibration, KALM encourages the team to improve existing practices rather than abandon them, which often preserves institutional knowledge and reduces disruption.
When to use KALM
KALM is ideal for mature teams that have established working patterns and need to fine-tune them. It works best when the team is not in crisis but feels that their ways of working could be more efficient. Think of it as a performance tuning tool rather than a firefighting tool. It is well-suited for quarterly reviews or whenever the team senses that the current balance of activities is slightly off.
Teams of four to ten people benefit most, and the format fits well into a 45 to 60 minute session. It is especially valuable for teams that run many ceremonies and processes—daily standups, sprint planning, backlog refinement, demos, retros—and need to evaluate whether the current mix is optimal. The "Less" column gives permission to reduce meeting frequency or duration without eliminating meetings entirely.
Avoid KALM when the team is genuinely doing things that should stop completely. If there are practices with zero value, KALM softness around elimination can be a liability. In those situations, DAKI or Start-Stop-Continue with their explicit "Drop" or "Stop" columns are more appropriate. Also avoid it for new teams that do not yet have established practices to calibrate.
How to facilitate KALM
Start by listing the team's current practices on a shared board: ceremonies, communication channels, tools, workflows, and agreements. This inventory provides concrete raw material for the discussion. Give the team six to eight minutes to write cards for all four columns, referencing the practice inventory as inspiration.
Process columns in this order: Keep first to anchor what is working, then More to identify underinvestment, then Less to find overinvestment, and finally Add to introduce new ideas. For the More and Less columns, ask participants to be specific about the adjustment: "More code review" is vague, but "more async code review and less synchronous walk-through" is actionable.
The key facilitation challenge is helping the team distinguish between Less and potential Stops. If someone writes "less of the weekly planning meeting" but means "this meeting is worthless," help them clarify: "Would reducing from 60 minutes to 30 minutes solve the problem, or does the meeting need to fundamentally change?" Sometimes Less is the right answer; sometimes the team is being polite about what should be a Stop. Close with commitments that specify the calibration: "Reduce standup from 15 minutes to 10 by focusing on blockers only."
Tips for getting the most out of KALM
The most powerful use of KALM is tracking calibrations over multiple retros. If the team says "more pair programming" one quarter and "less pair programming" the next, that oscillation reveals that the right level has not been found. Plot these shifts over time to identify practices that need a stable, well-defined standard rather than ad hoc adjustment.
Avoid letting "Add" become a wish list disconnected from "Less." Every addition creates demand on team capacity. When someone proposes adding a new practice, ask: "What would we do less of to create space for this?" This zero-sum framing ensures that additions are realistic and that the team is not inadvertently increasing their total process load.
Use the Keep column as a team constitution. Over multiple KALM sessions, items that consistently appear in Keep represent the team's core identity—the non-negotiable practices that define how they work. Make these explicit in a team charter document. When new members join, sharing the Keep list accelerates onboarding and sets clear expectations about team norms.
Variations and adaptations
For remote teams, KALM works exceptionally well asynchronously because the calibration decisions are specific enough to discuss in writing. Share the practice inventory and the four columns, give the team 24 hours to populate the board, and then hold a 30-minute synchronous session to vote and commit. Remote teams particularly benefit from the "Less" column for meeting-heavy organizations—it creates a constructive way to reduce meeting overhead.
For larger groups, run KALM as a survey before the retro. Ask each team member to rate current practices on a five-point scale from "much less" to "much more." Compile the results into a heatmap that shows which practices the team agrees need adjustment and which are contentious. The synchronous session focuses on the contentious items where opinions diverge.
A useful variation for teams managing their time carefully adds a time dimension: estimate how many hours per sprint each practice currently costs and how many it should cost after adjustment. This quantified KALM makes trade-offs visible—adding a two-hour practice and reducing another by two hours is a neutral swap, but adding two hours with no reduction is a net increase that needs to come from somewhere. Another adaptation combines KALM with a team health check, mapping each health dimension (e.g., collaboration, pace, learning) to specific Keep/Add/Less/More actions.
Run Retrospectives in CodeKudu
CodeKudu includes dozens of retrospective board templates, anonymous feedback, AI summaries, and action items that sync to GitHub Issues, Jira, and Linear.
