Getting Better at Getting Better—How the After Action Review Really Works: An Interview with Marilyn Darling

from Leverage Points Issue 61

Copyright © 2005 Pegasus Communications, Inc. (www.pegasuscom.com). All rights reserved. No part of this article may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying and recording, without written permission from Pegasus Communications, Inc. If you wish to distribute copies of this article, please contact our Permissions Department at 781-398-9700 or permissions@pegasuscom.com.

In 1989 Marilyn Darling founded Signet Consulting Group (www.signetconsulting.com) to conduct research and consulting in strategies for corporate learning. With her business partner, Charles Parry, she has evolved a practice called "emergent learning," that is, learning about your own work in the course of doing your work. About eight years ago, Marilyn was introduced to the After Action Review (AAR), the U.S. Army's learning practice that allows soldiers to extract lessons from one situation and apply them to another. Recognizing the AAR as "emergent learning on the ground," she has since incorporated the method into helping individuals and organizations raise the level of learning in their own work processes.

At the 2005 Pegasus Conference, Marilyn and Lieutenant Colonel Mark Pires (ret.) will be conducting a one-day pre-conference session, in which participants will learn and apply the tools of the AAR process to their work challenges (learn more about the conference). In the following interview, conducted by Leverage Points editor Kali Saposnick, Marilyn describes the challenges and benefits of transferring what we learn from one project to the next and how AARs empower that process.

Leverage Points: What is the After Action Review process, and why would organizations benefit from using it in their work?

Marilyn Darling: An After Action Review (AAR) is a tool for continually improving your results by discovering and applying lessons before, during, and after a project, and for applying those lessons to similar projects in the future. Many people believe that the main purpose of AARs is to capture lessons for the benefit of other teams. But our belief is that the team itself is the first, best customer for what is learning, and the best time to apply "lessons learned" is in the current project itself. What a shame to wait until the end of a project to hold an AAR and gain an insight that might have helped improve the results of that project!

LP: What is the first step in using AARs?

MD: You first have to figure out what part of your business you want to improve. One of the key reasons people misunderstand the After Action Review is because its name is misleading—it seems to assume that you do your learning after action is completed. In fact, you really learn when, at the beginning of a piece of action, you create a plan for testing out an idea and seeing if it works.

The real AAR process is a cycle, what we call the arc of learning. It involves a leader and his or her team asking at various stages of a project: What is the work that we do that we need to improve in order to produce the kind of results we want? Each time we begin that work, can we create a hypothesis about what will make us successful this time? Can we collectively try out this hypothesis and do an AAR at the end to ask: What was our intent? Did we actually accomplish it? What caused our results? What can we sustain? What can we improve?

I like to think of the AAR as a learning experiment in which you try to compare what you think you will accomplish with what you actually accomplish, from the beginning of the project to the end. As you repeat this experiment with each new phase of the project, you become more rigorous in your thinking rather than just identifying fixes. If you wait until the end of a project to ask these questions, you end up doing a postmortem, not a true learning practice. That's why you need to set the stage at the beginning, by establishing both the task and purpose for every piece of action, for every level, for all the different groups of people who are involved in enacting that piece of work.

LP: Can you give me an example of how companies are using the After Action Review?

MD: At Harley-Davidson, a leader at the Kansas City plant used an AAR practice to prepare his people for manufacturing new motorcycle models. He applied AARs to the "pre-build" process (in which you manufacture the product for a day to test out assumptions) to ensure that his team learned what it needed to build that new product to standard on day one. Before each pre-build, the team did a planning exercise, asking questions such as "What will it take for us to reach this level of precision and build this many bikes in this amount of time?" After a day of running the manufacturing line, they conducted a series of AARs in small task-focused teams, in which actual performance was matched against initial assumptions. After each AAR, assumptions were refined, standards were raised, and another pre-build was conducted.

Not only did the AAR practice produce performance improvements, it offered the bonus of increased team knowledge and confidence during production planning. Further, workers became excited about their increased knowledge of the whole operation and gained strong planning and data-gathering skills.

LP: Does the AAR work as well for functions such as sales and marketing, where outcomes are less tangible than product development?

MD: People in those functions may not be producing an actual product, but they do have "punctuated" experiences. By punctuated, I mean the defining moments, the things you repeat in your work, such as attending sales meetings or writing marketing proposals. Once you've identified those punctuated experiences, you can incorporate the AAR into them.

When I facilitate AARs, I use a timeline, which allows the group to pull out repeated tasks more easily. If you're focusing on sales that didn't succeed, rather than just listing what didn't work or your rationalization for why you didn't get the contract, look forward. Before the next call to a prospective client, ask: What worked last time? What didn't work? What got you closer to the sale? What got you further away? After you do this for a while, you will have a surprising amount of data to draw from. The more you begin to see the subsets of tasks you do as part of more complex, less tangible work, the easier it becomes to build learning into them.

LP: Many people would claim they do make a hypothesis at the beginning of a project and review their results afterwards—but they continue to repeat mistakes. Why don't the learnings get transferred?

MD: One difference between successful practices we've experienced and those that fail is the answer to the question: Who's the customer for the AAR? If you're documenting what you've learned in order to disseminate it to some other team, you're less likely to be successful than if you're applying it to your own work. If your team can work together again and apply your lessons learned to the next piece of action or the next project, then you have the opportunity to imbed the sensibility of learning—the quality, the feeling, the excitement of actually being able to improve results through learning—in your own practice. That is where AARs really gain a foothold in an organization.

What most people don't realize is that the weak link in the learning cycle is not between planning and action. Many of us do spend a lot of time planning. It's not even between action and reflection. If we take the time, we can in fact reflect on our actions. The weak link is between reflection and planning—applying what we learned at the end of one project to the initial planning stages of the next project. You can address this weak link by bringing up lessons learned at the outset. It's a leadership act to say, "What did we learn from last time?" If a leader or members of a team do not ask that question, they're unlikely to apply past learning to creating future success.

Many of us have those very clear, repeated pieces of work that are important to us, such as making conference presentations or conducting a particular training program five times a year. It's natural to ask yourself what you learned last time and find ways to improve the next time. But when you're dealing with teams of people, longer projects, and less clear pieces of action, it gets harder to achieve that quality of learning. That's what the AAR is trying to do—help teams of people actualize that same sense of learning that comes more readily to individuals.

LP: One of the biggest challenges of integrating AARs into the workplace is getting key people on a project to participate. How might members of a team convince a leader of an AAR's usefulness?

MD: One great aspect of AARs is that they produce data about intended versus actual results and what is causing the gap. The AAR will help you get better at whatever it is you intend to measure. And if you're not getting better, the AAR helps you build evidence about what's getting in the way. So for change agents, it is quite a good way to demonstrate the need to address something at a more systemic level.

As an organizational learning tool, the AAR is very concrete. You don't have to believe in its value to see it. You just have to do it two or three times around one piece of repeating action. You can actually see yourself create a difference as a result of being conscious about your actions. That value added is the kind of proof that leaders love. And in the process of applying AARs to your real work, you're learning about trust, communication, and the value of seeing the larger system and looking for leveraging points in it.

LP: Why don't people build more rigorous accountability for learning into projects?

MD: Some people feel that the AAR process saddles them with having to make some decisions in advance about how they're going to do something. They feel they can't innovate because they've defined their project too precisely, and they want to have the ability to adjust because situations change—a hundred different things are going to happen between the plan and the actuality of it.

The preparation for the AAR focuses on the place where you need to get very specific and precise: task and purpose. "I will pin down what I promise I'm going to deliver. I will also pin down what I think it's going to take to deliver it as well as how we're going to coordinate different units that are involved in the delivery. But I also recognize that, along the way, many things will happen that we'll have to react to, and my initial plan might not work."

It's stunning how much flexibility the detailed planning of the U.S. Army's AAR process gives unit leaders—they're synchronized, they know what their interdependencies are beforehand, and they recognize that their plan may not work. They have established task and purpose. A unit might say: "My task is to take this hill in order to ensure that the enemy does not pass through John Wayne Pass." The real purpose is that the enemy does not pass through John Wayne Pass. The soldiers might not take the hill and instead do something entirely different. But they recognize that any change of plan is going to affect three other units who are all expecting them to be there, and they need to coordinate with those units to make sure they adjust their plans simultaneously.

So you're not pinning yourself down to a particular task; rather, you're committing to a particular purpose with the recognition that you have to coordinate with other people to get there, which is what learning in complex environments is all about. And having done all of this in advance transforms the AAR meeting: "We said that if we took this action in this situation, we would produce that result. Was our thinking accurate?" Rather than just making a list of corrections, the AAR has become a tool for training a group's thinking process—a true vehicle for team learning.