Summer is probably one of the few times of the year when the pace slows down enough to take a step back from day-to-day learning and development priorities. Between business requests, mandatory training, content updates, new tools to evaluate and reports to produce, there is rarely enough time to look at the entire learning ecosystem from a broader perspective.
This moment of reflection comes at a particularly interesting time for L&D. Artificial intelligence is accelerating content creation, new tools are entering learning tech stacks, adaptive learning is opening up new possibilities for personalization, and increasingly sophisticated data is making it possible to look beyond completion rates alone.
Summer can therefore be an opportunity to ask a simple question: do today’s training programs still reflect the way we want people to learn?
Rather than another article about the latest L&D trends, here is a summer workbook. No crosswords or impossible math problems, but ten exercises designed to look differently at training programs, tools and some well-established instructional design habits.
There is no need to question everything. The idea is simply to use this quieter period to experiment, observe and perhaps return in September with a few new perspectives.
EXERCISE 1 : REVISITING THE TRAINING CATALOG
A good place to start is by opening a few courses from the training catalog and trying to complete this sentence: “By the end of this training, the learner will be able to…”
The exercise sounds obvious, but it can quickly become more difficult. Many learning objectives are still built around verbs such as “know,” “discover,” “understand,” or “become aware of.” These formulations express an intention, but they say relatively little about what someone will actually be able to do differently after completing the training.
A management course might, for example, aim to develop the ability to conduct a difficult feedback conversation. Sales training might focus on improving opportunity qualification. An AI course could help employees learn how to select the right tool for a particular situation or identify information that should not be shared with an AI model.
This distinction then influences the entire learning design process. When the objective is primarily to communicate information, a document or resource may sometimes be enough. When the objective is to develop a skill, the learning experience will generally need to create more room for practice, decision-making, mistakes and feedback.
The summer exercise is therefore to select an existing course and try to redefine its objective around an observable capability.
EXERCISE 2 : ASKING WHETHER THE TRAINING COULD HAVE BEEN A DOCUMENT
Not every piece of information within an organization necessarily needs to become a training course.
An infrequently used procedure, a reference list or certain regulatory information may simply need to be easy to access when required. Yet having an LMS and authoring tools available can naturally lead organizations to turn many learning requests into e-learning modules.
An interesting exercise is to take an existing course and imagine replacing it with a perfectly structured document in which employees could instantly find the information they need through a simple search.
If this transformation makes little difference to their ability to perform their work, the content may be more useful as a resource than as a training experience.
Training becomes particularly valuable when the goal is to develop capabilities that cannot simply be found with Ctrl + F: analyzing a situation, making a decision, explaining a line of reasoning, conducting a conversation or applying knowledge in a new context.
This distinction becomes even more important with generative AI. Creating learning content is now significantly faster. But easier production does not necessarily mean organizations should create more modules. Instead, it can create more time to determine what genuinely needs to be learned and practiced.
The summer exercise: identify one training module that could become a resource and, conversely, one resource that could benefit from becoming a genuine learning experience.
EXERCISE 3 : PUTTING YOURSELF IN THE LEARNER’S SHOES
Another useful experiment is to complete one of your own training courses under conditions that are as close as possible to those experienced by employees.
Not during a morning entirely dedicated to learning, but between two meetings, with emails coming in, a few notifications appearing and only 30 to 60 minutes available.
The objective is not necessarily to evaluate the quality of the content. It is primarily to observe the rhythm of the experience. At what point does attention begin to decline? When does the temptation to speed up a video or skim through the content appear? How much time can pass without having to answer a question, think about a situation or actively use what has just been learned?
A 45-minute course does not necessarily represent 45 minutes of learning. It may contain a significant amount of passive content consumption and only a few moments during which the learner is genuinely active.
This is one of the reasons conversational learning plays an important role in Complement’s approach. The objective is not simply to replace a traditional video with an avatar, but to allow learners to interact with an AI tutor, ask questions, be challenged and receive contextualized responses as they progress.
The experience can therefore move closer to a genuine learning conversation rather than a succession of content to consume.
The exercise is to identify the longest period of passivity in a course and imagine an interaction that could replace it.
EXERCISE 4 : TESTING AI-POWERED TRAINING CREATION
Choosing a familiar topic and asking a generative AI tool to produce an entire training course is now a particularly revealing exercise.
Within minutes, it is possible to generate a structure, learning objectives, several chapters, examples and even an initial set of assessment questions. What would have required several hours of production only a few years ago can now be generated almost instantly.
But the exercise becomes truly interesting once the content has been produced.
The next step is to look at what is missing. Do the scenarios resemble situations that actually occur within the organization? Does the learner have to reason or simply read? What happens when a concept is not understood? Is there an opportunity to ask a question? Does the level of difficulty evolve according to the learner’s mastery? Does the feedback take the learner’s response into account?
AI is profoundly changing learning content production, but creating a course faster does not automatically create a better learning experience.
At Complement, AI is also used to accelerate certain stages of course creation. Existing resources, such as presentations, can be used as a starting point to build modules and generate assessments. But the role of AI does not end once the course has been created. It also operates during the learning experience through interactions with the tutor, questions, assessments and personalization.
The challenge therefore becomes less about how to produce more with AI and more about understanding what AI can now make possible pedagogically at scale.
EXERCISE 5 : RETHINKING ASSESSMENT
Reviewing a few questions from an existing quiz can also be revealing. For each question, the objective is simply to identify what is actually being assessed.
Asking learners to list the different stages of constructive feedback, for example, can verify whether they remember a particular framework. That is useful information, but it does not necessarily reveal how they would react when faced with an employee who challenges that feedback or becomes defensive.
Presenting this situation and asking learners how they would respond requires them to use their knowledge differently. Instead of simply recognizing or recalling information, they have to apply it.
This does not mean multiple-choice questions should disappear. They remain useful for many learning objectives. But they can be complemented by scenarios in which knowledge must be mobilized in a context closer to real working conditions.
This approach is also reflected in the assessments available through Complement. Learners can be questioned orally by their tutor and formulate answers in their own words. Scenarios can also confront them with decisions similar to those they may encounter in their professional environment.
Assessment then stops being merely a validation mechanism placed at the end of a chapter. Through the feedback that follows each response, it can become a learning moment in its own right.
For this exercise, selecting a few existing multiple-choice questions and trying to turn one of them into a realistic scenario is already enough to shift perspective.
EXERCISE 6 : REVIEWING THE L&D TECH STACK
LMS, LXP, authoring tools, content libraries, virtual classroom platforms, coaching solutions, skills management platforms, generative AI… The technological environment surrounding L&D has expanded considerably in recent years.
As solutions accumulate, it can be useful to return to a very simple question: what problem does each tool actually solve?
The exercise involves temporarily forgetting about features. An LMS may play an essential role in administration, enrollment and compliance tracking. An authoring tool may facilitate content production. A library may provide access to generic learning resources. A coaching solution may address a need for practice or individualized support.
Looking at the stack from this perspective can reveal that several solutions are addressing the same problem or, conversely, that an important part of the learning experience remains largely uncovered.
The emergence of new solutions does not necessarily mean replacing the entire existing infrastructure. Complement, for example, can integrate with an existing LMS through LTI. The LMS can therefore retain its role within the learning ecosystem while another technological layer enriches the learning experience itself.
The question becomes less “Should the LMS be replaced?” and more “What role should each tool play within the learning experience?”
EXERCISE 7 : IMAGINING A TRAINING COURSE WITHOUT POWERPOINT
Imagining having to train several hundred employees on a new skill without using a single slide is an excellent instructional design exercise.
Without a presentation to immediately structure the content, the starting point naturally changes. The experience could begin with a situation, a problem to solve or a decision to make. Knowledge can then be introduced when it becomes necessary before placing the learner in a new situation where that knowledge has to be applied.
The exercise reverses a deeply established instructional design habit. Instead of starting with “What do we need to say?”, the process begins with “What does the learner need to be able to do?”
This obviously does not mean PowerPoint should disappear. Organizations already have countless presentations containing valuable internal expertise, and Complement can use existing resources as a starting point to accelerate course creation.
But the learning material can become the starting point of the experience rather than the experience itself. Conversations with the tutor, questions, scenarios and assessments can then be built around that content.
The exercise can therefore be as simple as imagining the first ten minutes of a course without starting by creating a slide.
EXERCISE 8 : DESIGNING THE SAME LEARNING JOURNEY FOR TWO DIFFERENT PROFILES
Imagining two employees taking the same course highlights another historical limitation of digital learning.
The first has just joined the organization and knows almost nothing about the subject. The second has been in the role for five years and already understands most of the concepts, while still experiencing a few specific difficulties.
In many learning environments, both people nevertheless receive exactly the same content, in the same order, answer the same questions and spend approximately the same amount of time completing the course.
This standardization is not always an instructional choice. It is also the result of a historical constraint: personalizing a learning journey for every individual was extremely difficult to achieve at scale.
Adaptive learning and artificial intelligence are gradually changing this constraint. The level of explanation, interactions and feedback can evolve according to the learner’s responses and level of understanding.
This is one of the principles Complement is seeking to develop: maintaining the ability to train at scale without necessarily delivering an identical experience to every learner.
The exercise is therefore to identify a few points in an existing learning journey where the experience could evolve according to the learner’s level or the difficulties they encounter.
EXERCISE 9 : LOOKING BEYOND COMPLETION RATES
Completion rates remain particularly convenient metrics. They indicate how many people started and completed a course and provide valuable information when monitoring the rollout of a learning program.
But completion alone does not reveal what has actually been learned.
An interesting exercise is therefore to imagine, for a moment, a dashboard where the completion rate has disappeared. What other data could demonstrate the effectiveness of a training program?
Progress between the beginning and end of a course could be measured. Concepts that learners struggle with could be identified. Recurring mistakes, questions asked or the ability to solve particular scenarios could also provide useful information.
The interactions generated during learning are making this kind of analysis increasingly accessible. When learners ask questions, answer orally or complete assessments, it becomes possible to gain a more detailed understanding of their strengths and difficulties.
At Complement, data generated through interactions and assessments can notably help identify concepts that learners have not fully understood and provide L&D teams with richer information about the learning experience.
The dashboard can therefore gradually evolve from an administrative monitoring tool into a tool for understanding how learners are progressing.
EXERCISE 10 : IMAGINING THE L&D MANAGER’S ROLE IN 2030
To finish this summer workbook, it can be interesting to look beyond training programs and consider the profession itself.
Some of the tasks that currently take up significant amounts of L&D teams’ time can already be accelerated with artificial intelligence. Generating an initial course structure, producing assessment questions, translating content, transforming existing resources or summarizing certain results is becoming progressively easier.
The more interesting question is what becomes more valuable when these activities require less time.
Understanding business challenges, identifying critical skills, interviewing subject-matter experts, defining learning objectives, designing relevant scenarios, choosing the right level of difficulty, evaluating AI-generated content, interpreting learning data and measuring impact could all become increasingly important.
The expertise of an L&D manager does not disappear because AI can generate a quiz or turn a presentation into a module. Instead, it shifts toward more strategic and pedagogical decisions: what genuinely needs to be learned, how to create the right conditions for learning, and how to determine whether meaningful progress has actually taken place.
This principle also guides part of Complement’s development. Automating or accelerating certain production stages can give learning teams more time to focus on what technology cannot simply decide on their behalf: the relevance and quality of the learning experience.
ONE REFLECTION TO TAKE INTO SEPTEMBER
The purpose of this summer workbook is obviously not to return in September with a new LMS, a completely rebuilt catalog, fifteen new KPIs and an AI strategy to deploy before the first coffee of the morning.
Instead, the idea is to identify one area worth exploring.
It could be making an overly passive course more interactive, bringing an assessment closer to real working situations, introducing more personalization into a standardized learning journey, or questioning a metric that ultimately reveals very little about actual learning. It could also mean reclaiming a few hours currently spent producing content and reinvesting them in instructional design.
At Complement, we start from a simple belief: technology becomes particularly valuable when it does more than help organizations produce and distribute more content. Its real potential lies in making some of the qualities of individualized support available at scale. Being able to ask a question, be challenged, receive feedback and progress differently according to what has already been mastered can fundamentally change the digital learning experience.
September will undoubtedly bring another wave of new tools, AI features and promises about the transformation of L&D. In that context, the role of the L&D manager may be precisely not to start with the technology.
But with a much simpler question: how can we create the best possible conditions for learning?
And there is no need to wait until September to start thinking about that.
Enjoy the summer.