Lessons learned: definition

Lessons learned is the structured practice of analyzing a project, experiment or incident once it is over, in order to extract explicit, reusable knowledge. The goal is not to judge people but to enrich the team's collective knowledge: what worked, what failed, and what the team would do differently next time.

Why lessons learned matter

A team that never formalizes its lessons learns half as much as it thinks: each member draws private conclusions that stay in their head and leave with them. The article Why startups forget their decisions shows the decision-side version of this problem; lessons learned are the learning-side counterpart.

Done well, the practice prevents the same mistakes from being replayed, speeds up onboarding (new hires inherit the lessons, not just the rules), and turns failures into assets: a failed test that produced a clear lesson generated value.

What lessons learned are not

  • Not a project report. A report tells what happened; lessons learned extract what should be remembered and changed.
  • Not a blame exercise. A session that hands out blame produces defensiveness, not learning. The golden rule: analyze causes, not people.
  • Not reserved for failures. Successes deserve the same analysis: knowing why something worked is the precondition for repeating it.

The practice also differs from its closest cousin, the post-mortem: a post-mortem is a lessons-learned format focused on a specific incident or failure, while lessons learned can cover any project, including successful ones.

How to run it in practice

A useful lessons-learned session has four steps:

  1. Reconstruct the facts: the timeline, the decisions made, the outcomes. Without shared facts, the analysis drifts into impressions.
  2. Analyze the gaps: what happened differently from the plan, and why? This is where the lessons come from.
  3. Phrase lessons as actions: not "communicate better," but "validate the brief with the client before production starts." A lesson that is not actionable is an opinion.
  4. Add them to the shared memory: a lesson that stays inside the session document is lost. It must join the team's memory and resurface at the next similar project. The post-mortem template structures the first three steps; for decisions that come out of the session, the decision record template takes over.

The weak point of the classic practice is step four: lessons get written, then never reread. A living memory like Verbasil addresses this by linking lessons, decisions and signals, and surfacing them when a similar topic comes back.

FAQ

What is the difference between lessons learned and a post-mortem?

A post-mortem is a specific lessons-learned format, focused on one incident or failure and run shortly after the event. Lessons learned is the general practice: it also applies to successful projects and routine experiments, on a rhythm less tied to urgency.

When should a team run a lessons-learned session?

At the end of any project or experiment the team wants to learn from: a launch, a channel test, a redesign, an incident. The right moment is close to the end of the episode, while the facts are fresh, but after things have calmed down, so that analysis wins over emotion.

Lire cet article en français