top of page

Deriving Lessons in the IDF - Taking it to the Next Level

1 October 2010
Dr. Moria Levy


As these lines are being written, fighting is taking place in Gaza. Alongside the fighting, the media is engaged in a debate and comparison of the operation to the Second Lebanon War. The word "lesson" is repeated many times in this debate. Even Nasrallah chose, in his televised speech from December 31, 2008, to refer to the IDF's derivation of lessons and its current significance in the fighting in Gaza. "The Israelis talked a lot about deriving lessons from the Lebanon War, but so far Olmert, Livni, and Barak have not announced what their goal is in the Gaza War," he was quoted as saying. This is not a positive reference, but even Nasrallah understands the importance of the IDF deriving lessons and tries to criticize it with this statement. The IDF's preoccupation with deriving lessons goes beyond the contexts of the wars in Lebanon and Gaza. The Google search engine counts about 88,700 mentions of the phrases "IDF" and "lessons." There is no doubt that deriving lessons is an important activity. Its importance stems from the aspiration to replicate successes, from wasting resources by "reinventing the wheel," and primarily from the need not to repeat mistakes. This need adds significance to the military's engagement with state security and human lives, where the cost of error is unbearable.

The purpose of this article is to propose an improved model for deriving lessons. The currently accepted conception includes two main stages, as described in the following diagram:

The proposed improvements are at the derivation process, at the level of the lessons themselves, and primarily in increasing the likelihood of future use of the derived lessons.

The IDF is already engaged in deriving lessons and learning from its experience. It intends to propose tools for taking this to the next level, improving the army's capabilities and performance.

The model is based on doctoral research work on this topic, currently being carried out by the author of the article, and on activities in the field that have been and are being done in other organizations outside the IDF, in R&D activities (two industrial organizations), on financial issues (two banks), on human resources issues and safety issues.

From Deriving Lessons to Lessons Management: The First Step Up

The term "deriving lessons" deals with developing new knowledge and understanding based on experience. We derive lessons not to torment ourselves or blame others but rather to ensure that the next time, we will do things correctly, well, and efficiently, based on what we know now after the previous experience; that we will repeat activities that contributed to success, and avoid activities that hindered success, or unfortunately, led to failure.

A lesson can be defined as "a recommendation for future conduct based on experience." This definition sharpens the dictionary's (Eben-Shoshan) definition of "a moral, a conclusion learned from experience," emphasizing the future, not just the past and present.

Learning is not instruction. We send our children to school not only to hear the teachers speak and instruct them but also so that they will know more and act according to what they have been taught.

In 1958, Jarvis defined learning as "a shift in performance when the stimulus situation and motivation remain essentially the same." Hence, internalization is required for learning to take place. Knowledge creation is not enough nor transmitted; it must be internalized and acted upon.

Accordingly, using the discipline’s name in Hebrew, "deriving lessons," is misleading. Deriving lessons emphasizes the creation of new knowledge and arriving at the Lesson. We require more than that. We seek the "lesson learned"- the lesson internalized in the new practice that creates a change in behavior and performance, enabling personal and organizational performance improvement.

It is, therefore, appropriate to use a broader term. Instead of "deriving lessons," it is proposed to use "lessons management." Lesson management deals with the lesson in a wider context and suits the need we seek to achieve - a "shift in performance."

We have expanded on the terminology. Of course, the main thing is not the terms themselves but what lies behind them. The wisdom in using the new terminology, "lessons management," is to create, by this use, tools that practically enable improvement in deriving lessons, improvement of the lessons themselves, and improvement in the level of use of the derived knowledge.

Deriving lessons alone is not sufficient. Studies show that all too often, we encounter situations where organizational employees invested thought, time, and effort, derived lessons, the lessons were presented to a relevant group of people, and conclusions (and presentations) were published, but when these or other employees needed to use the knowledge:

They did not remember what had been derived. We are inundated with information and knowledge, and the brain protects us by filtering out much of the knowledge or transferring it to back memory rooms, so we do not always remember the lessons we have already derived.

They did not connect the new situation to previous situations and did not use the lessons learned in other contexts. Our brain works through patterns. At any given moment, we perceive many cues and need to respond to them according to our knowledge. We try to fit the situation into an existing pattern and choose our response accordingly.

They did not know about the knowledge that had been derived. They arrived at the assignment and received a handover, but the person who came before them did not remember to share with them what had happened to their predecessor in the role since there was so much to share. They emphasized things they had experienced themselves or seemed most current at the time of the handover.

At first reading, some may think that such a situation is impossible. Every lesson they will argue is embedded in a procedure and becomes part of tomorrow's doctrine and conduct from today's paper. However, in practice, this is not the case due to several factors:

First, not every lesson finds its place in an appropriate procedure (we do not have a proper method for every situation and wise statement, and rightly so).

Second, because the skill of writing a lesson is a discipline in its own right, and guidance on how to do it properly is needed. The generalization process, for example, is essential in refining a lesson into a quality lesson. To what extent are people aware of this and implement it in practice? To what extent do people leave the specific context in which the lesson was derived and examine the broader contexts (other sectors, different types of activities, other technologies, etc.) to understand whether the lessons are also relevant in the additional contexts? Some may think this is a skill related to deriving lessons. Work experience with several organizations shows that many people can derive lessons, but refining these lessons into quality lessons is a separate task that requires skill. Attempting to impart this skill to all lesson-deriving personnel is not simple at all, and it is found that it is preferable to have the refining process carried out by a central entity, which can also perform additional processing tasks that require a comprehensive systems view. Writing a lesson is not an art, but it is certainly a craft.

The proposed model argues that four main stages together make up the life cycle of lesson management:

  1. The lesson derivation stage

  2. The lesson repository management stage

  3. The lesson dissemination stage.

  4. The stage of promoting the use of lessons

These stages can be presented in the following diagram:

In defining a complete life cycle, emphasis is given not only to creating knowledge but also to its refinement (into higher quality knowledge), making it accessible to the employee, and assisting in bringing them closer to it through a variety of tools (the stage of promoting the use of lessons). The complete life cycle that includes all these stages, and at its end, the promotion of use as an independent and separate stage, whose importance and place are no less than the production stage, seems to assist in increasing the chances of using lessons in organizations that have experienced and are experiencing this model.

It is worth explaining the choice of the name for the final stage: "promoting use." Just as we cannot force-feed the cow but only lead it to the trough, we cannot consistently enforce the use of lessons in every situation. Sometimes, we will enforce them, and sometimes, we will bring them to the doorstep of the commander or soldier who needs to use them. Hence the cautious choice of name for this stage: "promoting use." Often, we will succeed in leading to actual use, but we cannot guarantee it; we only guide towards that point by promoting use.

Non-Binding Lessons: A Second-Level Upgrade

In defining the word "lessons," we use the word "recommendation": "A recommendation for future conduct based on experience." Using this word is unclear or provokes opposition among some readers. Why a recommendation and not an obligation? If we have experienced, erred, or succeeded, why not obligate others based on this as well?

It is correct to clarify. The proposed model is okay with binding lessons. Some lessons are indeed appropriate to define as binding. It is necessary to describe them as a procedure and define appropriate punishment for those who do not comply. However, it should be recognized that dealing only with binding lessons dramatically limits the scope of lessons that can be derived and certainly limits the scope of learning. For every definite lesson, always correct, many recommendations and emphases can add significantly to the knowledge and performance of those who come after us. Recommendations that will not be enforced in some cases but have a rationale; recommendations that led to success, but we are not clear why; emphases for consideration when acting. Obligating the user to act upon them is not always appropriate since we do not have enough tools to assess all the other field conditions. The recommendation and emphasis allow us to bring the knowledge to the commander's doorstep and leave the discretion in their hands, according to the specific case and its characteristics.

This simple extension can significantly expand the space of lessons and the professionalism of the people. Most experts' knowledge is in the form of recommendations, not necessarily absolute truths. The extension allows for more significant learning and professionalism than a state that defines lessons only as binding.

Experience Alongside Lessons: A Third-Level Upgrade

Just as broadening the reference to a lesson to its original meaning (which is not necessarily binding) increases the potential scope of learning, so does dealing with experience alongside lessons. Experience, in English, is the knowledge accumulated by every commander and soldier during their activities. Unlike lessons, which are derived through an active process in which the organization stops its work and creates, through more or less structured thought processes, new knowledge and experience accumulate even without active thought processes aimed at that purpose.

One of the most significant books in the classics of knowledge development and adult learning is Kolb's 1984 book Experiential Learning: Experience as the Source of Learning and Development." In this book, which deals with learning styles, Kolb emphasizes the importance of learning from experience. Everyday observation teaches that this is true at all different age levels. Infants, for example, learn to eat and walk through experience. They improve these skills based on experience.

Those who say we should not turn all learning processes into one lump would be right. There are many ways of learning, and deriving lessons and experience are just two cases. It is right to look at these two learning processes in combination since the products of deriving lessons are like the products of experience. When we benefit from the combination, we should maintain it. This is the case with lessons and experience: life experience proves that we can hardly distinguish from the wording of the knowledge before us, whether obtained from an orderly lesson-learning process or intuitive experience.

The proposed model argues that it does not matter at all. Suppose the knowledge created is valid and verified. In that case, we believe that it can assist us in the future, and there is no reason to ponder its origin and distinguish between "lesson" knowledge and "experience" knowledge. It can be estimated that the combination of the experience existing in the consciousness of the knowledge holders in creating the lessons they conduct occurs in any case – consciously or unconsciously.

The experience existing in the minds of commanders and soldiers is an indescribable organizational asset. Collecting it in a focused manner and adding it to the base of lessons can significantly enrich it.

The word "insight" generalizes experience and lessons into a joint family. Insight is a piece of knowledge. Its source can be a lesson (an active creation process) or experience (a creation process during the activity).

This generalization brings us to an extended model of the lesson management life cycle, as described in the following diagram:

Managing an Insights Repository: A Fourth-Level Upgrade

For years, we have become accustomed to managing investigation documents in libraries. Despite the technological improvements over the years, libraries are repositories that are not easily accessible enough. Imagine a brigade commander preparing for an operation. Assume even that he has time. He enters the library and begins to peruse one investigation document. He searches for the context and reaches the lessons. From there, to the second document and the third. Very quickly, he begins to scan partial documents, glance, and search only for lessons, often embedded somewhere within the document. The investigations may also include lessons that repeat lessons that have already been derived, in the best case, and lessons that contradict previous lessons he has read in the less good case. It is an ineffective reading and study, and in most cases, even if that brigade commander has time (and this point is undoubtedly questionable), the chances are slim that he will manage to get through the fifth or sixth document.

Lessons and experience are pieces of knowledge. It would have been preferable for that brigade commander to be exposed only to these pieces of knowledge, which we have termed insights, and not to the entire investigation that led to these conclusions and lessons. It is essential to allow for an in-depth study of the inquiry itself, if necessary, but first, without overwhelming the reader with information.

The model proposes managing the insights (lessons and experience) in a repository. A repository that includes the pieces of knowledge, access to the sources they rely on (links to documents), and attributes. Attributes are an essential part of defining any repository. If we have an object, we can say that one of its attributes is its color (values: red, green, yellow...); another attribute is its size (values: small, medium, large), and a third attribute could be its shape (values: rectangular, round...). Adding the attributes and their values is a tool that helps filter the information presented to the reader. Instead of the reader having to read all the insights appearing in the repository, they can filter those relevant to the case they are currently facing and focus on the knowledge that will help them most in performing their task. In professional language, we refer to the attributes as the contexts in which we think the lessons are relevant for the future. According to the nature of the repository (operational, technological, economic, laboratory, logistical, etc.), appropriate attributes will be defined (respectively: sector, type of activity, components, discipline, process, etc.). The importance of the attributes is in making the information accessible to the reader. Information overload, we must remember, is almost equivalent to a lack of information. The bottom line in both is the non-use of knowledge.

There is no expectation for every commander investigating the field to extract all the lessons and characterize them. This requires skill, but more than that, it requires a systems view. A systems view allows one to see the contexts in which the knowledge was born and additional contexts of when else the knowledge is valid and applicable. Such a view can be done by a common integrating entity at the unit/corps/arm level, an insights repository manager, whose roles include:

  1. Assisting in structuring the initial repository.

  2. Refining every new insight that arrives:

    1. Including it in the cases and contexts in which it is relevant.

    2. Phrasing it in an agreed upon, practical, and value-added way.

    3. Merging it with the other existing insights to examine repetitions, complementarities, and contradictions and resolve them.

  3. Maintaining the existing: monthly/quarterly examination of outdated insights to see if they are still valid, require refreshing or can be removed altogether.

This is not meant to fill a position or create a new role. From experience working in several organizations, this is a role that, after the initial establishment, does not consume more than low percentages of a position. Although we are always short on roles, people, and time, this is time worth devoting since the lessons are the "cream" of the organizational knowledge, and it is not sure that organizations, in general, and the IDF, in particular, can afford to forgo the advantage of managing this core asset.

The repository, as defined, is quite accessible and is found to have several advantages:

  1. Shortening the time to reach the critical knowledge contained in the insights (thanks to the management of single pieces of knowledge and thanks to attributes).

  2. Increasing the number of contexts for each insight is relevant (exiting the event context and moving to contexts where the knowledge is valid/appropriate).

  3. Increasing the chances of accessing the repository (due to the high accessibility the repository provides).

This repository is undoubtedly an upgrade that meets most of the needs defined in the article's goals.

Wide Dissemination of Insights: A Fifth-Level Upgrade

The IDF is a large organization with multiple arms, corps, and units. The commendable lessons production culture in the IDF creates a situation where new knowledge is produced every day in some units of the IDF. Managing lessons in each unit separately cannot utilize the insights created entirely. The knowledge dissemination mechanism should be planned intelligently: Initially, some of it is local and relevant only to very focused cases. However, many lessons could be suitable for additional units, parallel or similar bodies in different commands and arms. A lack of policy leads to situations in large organizations where non-critical lessons remain within the unit's boundaries and are not disseminated to other units despite their potential for sharing and relevance to the different units. Planning a dissemination mechanism will allow and encourage dissemination but prevent overload (excess resulting from everyone disseminating everything to everyone). Such a mechanism can rely on three essential bases:

Clear dissemination rules

Defined repository structure and attributes that allow for swift and efficient transfer and filtering.

Repository managers who are aware of parallel and complementary repositories know them and can, therefore, exercise additional discretion and transfer knowledge only to those they deem to have the potential to benefit from it and use it.

Promoting the Use of Insights: A Sixth-Level Upgrade

So far, five improvement components have been described, which can primarily assist in improving the quality and scope of lessons and increasing their accessibility.

Promoting usage is the process of translating new knowledge into action. Its absence leads to a situation akin to product development, which neglects marketing and sales efforts.

Until today, most efforts in most organizations have focused on instilling lessons through two channels:

  1. Immediate, one-time notification about the new lessons will be sent to all relevant role-holders.

  2. Anchoring the lessons in procedures.

In most organizations, the use of these two channels alone is insufficient since each one has an inherent problem:
  1. One-time notifications lose effectiveness since people change roles. Who was there yesterday may not be there tomorrow. In the IDF, this phenomenon is exacerbated by short tenures and sector rotations. We intend to notify role-holders, but in practice, we notify people filling those roles. The people change, and some information is lost during the role handover.

  2. Procedures are limited on two levels:

    1. Procedures are limited in scope. Not every derived lesson is suitable for an existing procedure, and not every suitable lesson is appropriate. Is it worthwhile to develop a new method where none exists? This is not to diminish the potential utility inherent in the lessons. Many small lessons that see wide use can significantly impact their overall utility.

    2. The phenomenon of procedures not being implemented is widespread in organizations and is exposed during quality audits or when failures occur. It turns out that it is challenging to enforce procedures, even on critical issues like safety. Non-implementation of procedures is especially common when dealing with edge cases that are not routinely practiced. A lack of awareness of the very existence of the procedure characterizes these situations.

Therefore, additional channels must be developed to promote usage. The model, not coincidentally, does not use the term "instilling." Instilling is portrayed as a one-time process, a stage in an activity with a beginning and an end. One cannot instill in a role-holder who has not yet been recruited. The term "promoting usage" includes one-time aspects, as in instilling, and permanent aspects, as part of routine activity.

Channels for promoting usage could include:

  • Work processes: Integrating the insights repository into routine work processes. For example, Before any operation, the commander should raise three insights he read in the repository that seem relevant to the planned activity as part of the operations plan submitted for approval. Making the repository accessible as knowledge nuggets accompanied by attributes makes this recommendation feasible since access does not consume too much time.

  • Training: Integrating the insights into the training framework. Familiarizing with relevant insights repositories in training frameworks, using them as part of the routine training framework, and integrating the use of insights repositories into training missions and lectures on any topic. Each lecturer can be provided with the relevant insights from the repository along with the lecture schedule. This way, a double benefit can be achieved: both usage and validation for insights that may no longer be accurate or correct.

  • Drills: Incorporating insights into the drill framework. This includes insights that will be drilled into every exercise program.

  • Portals: Integrating insights repositories into knowledge sites and portals dealing with professional topics related to the repository. The integration can be inclusive (referring to the repository and increasing its availability to the user) or focused (on a specific information page, exposing lessons on a defined sub-topic).

  • Information systems: The insights repositories must be accessible from the routine computerized work environment, whether through dedicated information systems or from the desktop in general.

More channels for promoting usage can be listed. Initial examples are given here, and each commander can choose and adapt the usage promotion channels to the work environment in which the specific body operates under its constraints.

As mentioned, promoting usage is one of the most critical components for increasing the chances of using the accumulated knowledge. These steps are sometimes relatively simple, yet their usefulness is high.


Fully utilizing lessons and all the essential knowledge accumulated in an organization is not as easy as we would like. This article presented various challenges, proposed steps, and levels, each allowing an improvement in achieving the defined goals: a little regarding the production process and much regarding the number of lessons, their quality, and the level of use. The discussion on optimal lesson production processes is interesting, but it is beyond the scope of this article. The discourse was expanded from lessons to insights, which generalize experience and lessons together, and a complete life cycle for managing these insights was proposed, starting from the production stage through repository management, dissemination, and finally, promoting usage. Implementing this life cycle significantly increases the number of times the organization produces lessons. In many organizations, the feeling of producing lessons without appropriate utilization in scope has weakened the hands of the producers. In cases where people experience repeated use, creating additional new lessons is easier.

Despite using the term "levels," most can be implemented in parallel. Each commander can choose which levels are more relevant to the body in which they operate and implement them even without full implementation. This is even recommended. Even if all changes are appropriate and reasonable, proper change management is carried out gradually and not all at once. The order can also be changed. The order presented explains the model and is not necessarily the order of implementation when an organization comes to change and improve.

The work is excellent, but the goal is clear: to turn a lesson from a recommendation or instruction into knowledge and actions; to create learning in its original meaning - Shift in Performance; to turn a Lesson into a Lesson Learned and thus improve the organization's performance.

A person in military uniform holding a computer
bottom of page