Action Learning

Action Learning is a workshop design approach in which participants work on their own real, unresolved problems over time, take action between sessions, and return to the group to reflect on what happened. The problem is not simulated. The action is not hypothetical. The learning comes from the cycle of committing to action, taking it in the participant's own work, and examining what the action produced. The method requires multiple sessions spanning weeks or months — one-session formats cannot produce what Action Learning is designed to produce.

The approach was developed by Reg Revans. As a physicist turned management researcher, Revans worked with Welsh coal miners in the 1940s and later with managers at the National Coal Board in the 1950s. He argued that managers facing genuinely new problems learn better by working on those problems together than by receiving programmed instruction. His compact formula — L = P + Q, learning equals programmed knowledge plus questioning insight — became the foundation of the method. He consolidated his approach in ABC of Action Learning in 1983.

Mike Pedler and colleagues carried the method forward in the UK management-education tradition. Action Learning for Managers (2008) documents how the method adapts to contemporary professional development — the structured set, the set adviser's role, the discipline of questioning over advice-giving, and the protocols that keep the set productive.

Michael Marquardt's Optimizing the Power of Action Learning (2011) documents the method's spread across global executive development, leadership training, and organizational learning. The method has been adapted for contexts Revans did not originally design for, including contemporary professional cohorts and creator-economy programs that run over weeks or months.

Action Learning closes the Active Learning series because it extends what every prior method does into the participant's ongoing work. Problem-Based Learning solves a problem inside the session; Action Learning works a problem across sessions. Case-Based Learning analyzes a resolved case; Action Learning works a case the participant is still living. Collaborative Learning happens in real time in the room; Action Learning extends collaboration across the weeks between rooms. The method is the bridge from session to practice.

What Action Learning Is

Four structural components define Action Learning, whether participants run the method in a group or in an individual configuration.

• A real, live, unresolved problem. The problem is the participant's own — something they are currently facing in their work, where the outcome is not yet determined and their action will shape what happens next. Simulated problems produce simulated learning. Action Learning only works when the stakes are real for the participant.
• The action learning set. The set is the small, consistent group of peers that meets regularly — typically four to eight participants in the canonical group form. The set stays together across the whole program, which lets trust build and the peers learn each other's situations deeply enough to ask sharper questions each session.
• Questioning, not advice-giving. Peers support each other through probing questions rather than by offering solutions. Revans' L = P + Q formula captures it: the programmed knowledge each participant brings combines with the questioning insight the set produces. The set adviser trains the group out of advice mode — the discipline is hard to hold and easy to abandon.
• Action between meetings plus reflection on that action. Each session ends with each participant committing to a specific action they will take before the next meeting. The next session begins with each participant reporting on what they actually did and what it produced. The cycle of commit → act → reflect is where the learning lives — not in the meeting itself.

Group and Individual Configurations

Action Learning's canonical form is the group set — Revans' original method, the UK management-education tradition, the research base Pedler and Marquardt document. Workshop applications also run in individual configuration, where a single participant works with a coach or adviser who plays the set's role. Each produces a different kind of learning.

Group configuration. Four to eight participants form a set that meets regularly — typically every two to four weeks over a three- to twelve-month program. Each meeting rotates a "hot seat" where one participant presents their current problem, the set asks probing questions, and the participant commits to the next action. All participants report on their between-session action at each meeting. The set adviser trains and maintains the questioning discipline.

Benefits of the group form:

• Peer perspectives surface blind spots the participant cannot see in their own situation.
• Hearing other participants' problems builds pattern recognition across contexts.
• The social accountability of the set drives between-session action in a way self-directed programs rarely produce.
• The set becomes an ongoing resource that extends beyond the formal program.

Individual configuration. One participant works with a coach or adviser who plays the set's role — asking the probing questions, holding the questioning discipline, and receiving the reports on between-session action. This fits 1:1 coaching engagements, executive development relationships, or situations where the participant's context is specific enough that peer discussion would not add much.

Benefits of the individual form:

• The participant's own context gets full attention in every meeting.
• Cycles run faster — no scheduling coordination across multiple people, so meetings can happen weekly or on-demand.
• The coach-participant relationship deepens across the program, which sharpens the questioning.
• No averaging across peer situations — every question and every reflection is directed at the participant's specific work.

Choosing the configuration. Group is stronger when peer learning is part of the outcome, when cross-pollination across participants' contexts would sharpen the work, or when the social accountability of the set is a structural requirement. Individual is stronger when the context is very specific, when faster cycles matter, or when the participant needs deeper one-on-one attention than a set can provide. Hybrid configurations are common — a participant in 1:1 coaching may also join a set for the peer dimension, or a set program may pair each participant with an individual adviser between meetings.

When to Use It

Action Learning fits programs where the outcome is sustained behavior change in the participant's own work — programs where participants face real ongoing problems, where one-session intervention cannot produce what the program promises, and where the learning has to happen across time rather than inside a single session. It is strongest when the workshop leader can commit to a multi-session structure and when the participants can commit to taking real action between meetings.

Good candidates:

• Multi-session programs spanning weeks or months, with real ongoing work between sessions.
• Cohorts where every participant faces a real, unresolved problem in their actual work.
• Outcomes that require changed behavior in the participant's live context, not just new knowledge.
• Participants with enough authority over their own situation to actually take action.

Less suitable:

• One-off workshops where there is no "between sessions" for action to happen in.
• Programs where participants lack real, unresolved problems to work on.
• Audiences who expect direct instruction and will experience the questioning discipline as the leader withholding help.
• Contexts where the participant cannot take action in their own setting — no authority, no live work, no stakes.

What It Looks Like in Practice

A six-month group-configured Action Learning program for six independent coaches working on scaling their businesses might run like this:

• Program structure. Six coaches form a set that meets every two weeks for 90 minutes, across six months — twelve sessions total. Each meeting rotates a "hot seat" across the set so every participant presents at least twice over the program.
• Opening session (session 1). The set adviser introduces the method — real problems only, questioning not advice-giving, commitments to action between meetings, confidentiality within the set. Each coach names the problem they will work on across the program. The problems are genuine: one coach is trying to raise prices without losing clients; one is trying to hire a first team member; one is trying to redesign their group program after a failed launch.
• Typical session (sessions 2–11). One coach takes the hot seat. They present the current state of their problem — what they tried last time, what happened, where they are stuck now (10 minutes). The set asks probing questions for 45 minutes — what would you expect to see if your current approach were working? What evidence from the past two weeks did you discount? What are you optimizing for that you have not named? The coach commits to one specific action for the next two weeks (10 minutes). The remaining five coaches each report briefly on the action they took since the last session (20 minutes).
• Between sessions. Each coach takes the action they committed to. The action is real — they send the pricing email, they post the hiring listing, they rewrite the program outline. They take notes on what happened.
• Closing session (session 12). The set adviser runs a final reflection across the program. Each coach names what they learned about their problem and what they learned about how they work. The set names the patterns that appeared across all six participants.

An individual-configured version of the same program keeps the same cycle: real problem, commit → act → reflect, questioning discipline. The coach or adviser plays the set's role in every meeting, and the meetings can happen more frequently because there is no scheduling coordination across six people. The participant still commits to real action between meetings and still reports on what the action produced.

How It Differs From the Other Methods

• From Problem-Based Learning: Problem-Based Learning uses simulated or historical problems within a session; Action Learning uses the participant's own live, unresolved problem across sessions. Problem-Based works in one sitting; Action Learning spans weeks or months.
• From Project-Based Learning: Project-Based Learning produces a tangible artifact through the program; Action Learning produces behavior change in the participant's ongoing work. Both extend across time, but Project-Based is organized around the artifact, and Action Learning is organized around the action.
• From Inquiry-Based Learning: Inquiry-Based Learning can remain cognitive throughout — a participant can leave with understanding and no action taken. Action Learning requires live action between meetings as a structural element of the method.
• From Case-Based Learning: Case-Based Learning analyzes resolved scenarios retrospectively; Action Learning works on the participant's current unresolved situation prospectively. Case-based is someone else's story; Action Learning is the participant's own.
• From Collaborative Learning: Collaborative Learning happens in real time during the session; Action Learning extends collaboration across the weeks between sessions and includes the participant's individual action between meetings. The set is collaborative, but the action is individual.
• From Simulation-Based Learning: Action Learning takes real action in the real context, with real consequences. Simulation-Based Learning practices the action in a safe context, with consequences that are intentionally cheaper. Action happens in the participant's live work; simulation happens in the session's controlled space.

Design Considerations

Action Learning has specific failure modes the design has to account for.

• Choose participants with real, ongoing problems. The method cannot run on invented problems or on participants who do not have authority over their own situation. Before the program starts, confirm that every participant has a genuine, unresolved problem they can take action on.
• Commit to the multi-session structure. One-off workshops cannot produce what Action Learning produces. The program needs multiple sessions spread across weeks or months so the act-and-reflect cycle has time to run. If the workshop leader cannot commit to this structure, a different method fits better.
• Train the set out of advice-giving. The hardest move in Action Learning is holding the questioning discipline when participants are stuck and peers want to help. Advice-giving collapses the method back into traditional coaching. The set adviser names this explicitly at the start and reinforces it every session.
• Establish confidentiality. Participants will bring real, current problems — often ones they have not discussed outside the set. Without explicit confidentiality agreements, the method collapses into polite generality. The set adviser establishes the rules in the opening session.
• Make between-session action structurally required. Participants who arrive at the next meeting without having taken action have no material for the set to work on. The design has to make action a precondition for the next session — typically through explicit commitments recorded at the close of each meeting and reported on at the start of the next.
• Match the set size to the meeting length. In group configuration, four to eight participants is the workable range. With ninety minutes and six participants, each hot-seat session runs about an hour and the remaining participants each report for a few minutes. Sets larger than eight cannot give every participant meaningful time; sets smaller than four thin the peer perspective the method relies on.

Connection to the Guiding Principles

Action Learning activates all seven principles, with particular weight on focused outcomes, purposeful initiative, and honest reflection.

• Active Learning (80/20): Participants are working, taking action, and reflecting for the vast majority of the program; the workshop leader or set adviser's 20% is framing, holding the questioning discipline, and running the reflection. Most of the learning happens between sessions, in the participant's own work — the time split extends beyond the session itself.
• Autonomy: The participant owns the problem, the action, and the conclusions. The set adviser designs the structure and holds the method; the participant does the work in their own context. Each participant is the final authority on what they do next.
• Purposeful Initiative: A participant who does not take the committed action between meetings has nothing to report and cannot meaningfully engage in the next session. The natural consequence is immediate and visible to the whole set. This is one of the tightest accountability loops in any method.
• Engaged Interest: Working on your own real problem is engaging on its own. The participant cares because the stakes are actual — the outcome affects their actual work, not a simulation.
• Status Awareness: The set adviser reads where each participant is by listening to the quality of their commitments, the reports on action taken, and the questions the set asks. A participant who is stuck, over-committing, or avoiding the real problem is visible across sessions.
• Honest Reflection: Reflection is structurally built into every session — every participant reports on what they did and what it produced. Over twelve sessions, honest reflection becomes a habit, not a one-time exercise.
• Focused Outcomes: The problem each participant works on is chosen specifically because its resolution will produce the capability the program exists for. Price-raising problems build pricing capability; hiring problems build hiring capability. The outcome is what the action produces in the participant's real work.

Action Learning is particularly well-aligned with the EVALUATE phase of the Workshop Doctor Roadmap — the phase where the workshop leader cycles between taking action on their own workshop and reflecting on what the action produced. The act-and-reflect cycle that Action Learning builds into the participant's workplace practice is the same cycle the Roadmap's EVALUATE phase asks the workshop leader to run on their own design.

Where this goes next: Action Learning takes the practiced performance into live work. The next method is its practice counterpart — the one that rehearses the same kind of interaction in a safe context before it has to happen for real. See Simulation-Based Learning.

Related

Read this next:

• EVALUATE — Measure Learning, Not Happiness — the Roadmap phase where the act-and-reflect discipline Action Learning builds becomes the workshop leader's own practice

Foundations of this concept:

• What Active Learning Is (and What It Isn't) — the pillar page; the definition Action Learning operates on top of
• The Guiding Principles of Active Learning — the seven principles Action Learning activates with particular weight on focused outcomes and honest reflection
• Problem-Based Learning — the adjacent method; problem-based works a problem inside one session, action learning works one across sessions
• Collaborative Learning — the adjacent method; collaborative runs real-time in the room, action learning extends collaboration across the weeks between rooms

Where this leads:

• EVALUATE — Measure Learning, Not Happiness — where the participant's act-and-reflect cycle becomes the leader's evaluative discipline

References

Marquardt, M. J. (2011). Optimizing the power of action learning: Real-time strategies for developing leaders, building teams, and transforming organizations (2nd ed.). Nicholas Brealey.

Pedler, M. (2008). Action learning for managers (2nd ed.). Gower.

Revans, R. W. (1983). The ABC of action learning. Chartwell-Bratt.