4Ls and Mad Sad Glad: learning and feeling retros

Many teams run retrospectives that treat the previous Sprint as a series of events to be graded. That produces a list of process problems, which is useful, and misses a second kind of information: how the work felt, what people learned, and what they wished they had. Two common formats are built to collect it.

Where these formats stand

Neither format is defined by the Scrum Guide or the SBOK® Guide. The Guide's description of the Retrospective is deliberately broad: the Scrum Team inspects how the last Sprint went with regard to individuals, interactions, processes, tools and the Definition of Done, discusses what went well, what problems it met and how they were or were not solved, and identifies the most helpful changes. Individuals and interactions are part of it. Feelings and learning are therefore within scope. A vendor guide to retrospectives lists "Glad, Sad, Mad" among the prompts teams can use to vary the format, and suggests focusing on the sad and mad items so that, over time, the glad column fills up.

The 4Ls

The four prompts are:

Liked. What the team enjoyed or valued in the Sprint. It captures strengths, and gives a baseline of what to protect.

Learned. What people found out: a technique, a fact about the product, a mistake and its cause. It treats the Sprint as a learning cycle, which echoes the empirical idea that the Guide places at the base of Scrum.

Lacked. What was missing: information, a tool, support, time. It points to resources, which the Scrum Master or the organisation may be able to supply.

Longed for. What people wished for. It is the most open of the four and encourages ideas that do not fit under any other heading, such as "a quiet day without meetings" or "a way to see how customers use this".

The format is gentle and works well for teams that find "what went wrong" a threatening prompt.

Mad, Sad, Glad

Mad, Sad, Glad asks for emotions directly.

Mad. What frustrated or angered people: an interruption, an unfair decision, a repeated problem.

Sad. What disappointed them: a feature that was cut, a thing that could have gone better.

Glad. What pleased them: a good conversation, a hard problem solved, thanks from a stakeholder.

The vendor guide's advice is to concentrate on the sad and mad lists and ask how to improve, so that next time there are only items in the glad column. A good facilitator adds a step the board does not show, moving from the emotion to the cause and then to a change. "I was mad that the deployment failed twice" becomes "what would stop it failing twice?"

How the two compare

A worked board

This is an illustrative 4Ls board after a Sprint in which a team integrated with a new partner system.

In the discussion, the team sees that three cards, the sandbox, the missing contact and the error codes, point to one need: reliable information about the partner. They decide to ask the Product Owner to arrange a named contact, to write up what they have learned in a short page and to pair on a replay tool in the next Sprint. "A quiet day" goes to the Scrum Master to raise with the wider organisation.

Now a Mad, Sad, Glad version of the same Sprint might have yielded: Mad, the two failed deployments; Sad, a demo that went badly; Glad, the pairing. The first board finds the cause of the deployments in the lack of information. The second finds the feeling and then needs to ask why. Neither is better. They produce different material.

Choosing between them

A simple rule of thumb: use the 4Ls when the team wants to learn and to find out what is missing, and use Mad Sad Glad when something has been bothering people and the team needs to say so. If you are unsure, start with the 4Ls, which are easier on a new group, and keep Mad Sad Glad for a Sprint that was clearly hard.

A useful variation is to combine them: ask for the 4Ls first, then ask which of the items made people glad, sad or mad. The cross-over shows which learning or lack mattered emotionally, and those are the items the team is most likely to act on.

When feelings help

They help when morale is a real constraint, when a team is not saying what it feels, or when a Sprint was stressful in a way that process questions do not reach. A team that has been under pressure for months may need to say so before anything else will work. The Guide's value of openness and its statement that team members have the courage to do the right thing and to work on tough problems apply: naming a feeling is a small act of both.

When they get in the way

• Venting without change. If every Mad item is a complaint with no action, the format teaches people that talking is enough.
• Blame. Emotion cards sometimes name individuals. Set the rule at the start that cards describe situations and not people.
• Discomfort. Some people dislike talking about feelings at work. Allow anonymous cards, and do not insist on explanation.
• Over-reading. A single angry card is not a trend. Look for patterns across the team and across Sprints.
• Manager presence. If a manager attends, the openness of Mad cards will suffer. The Guide places the Retrospective among the Scrum Team.

See psychological safety in retrospectives for how to create the conditions in which these formats work.

How to practise it

Try the 4Ls once and compare the actions it produces with those from your usual format. If they differ, keep both in rotation. In the simulator the retrospective at step 4.2 prompts reflection on what to change, and you can bring the 4Ls into your own reasoning.