Why It Should Be Chosen Selectively
Learning in the flow of work has quickly become one of the most popular ideas in Learning and Development (L&D). And for good reason. It promises speed. Relevance. Minimal disruption. Support at the moment of need. In the right situations, it delivers all of that. But there is a growing problem.
Many organizations are not using learning in the flow of work as a targeted solution. They are using it as a default response. And in doing so, they are making the same mistake they made with courses—just with a different modality.
The Pattern We Have Seen Before
For years, the default answer to most performance problems was: “Build a course.” Now, in many organizations, that default is shifting to: “Put it in the flow of work.” At first glance, this feels like progress. It sounds more modern. More efficient. More aligned to how people actually work. But the underlying decision logic often hasn’t changed. The modality is still being chosen first. The problem is still being defined second. That is where things start to break down.
Modality Is Not A Strategy
Whether something is delivered as:
A course
A workshop
A job aid
A checklist
A prompt library
Or embedded directly into the workflow
…none of those choices are inherently right or wrong. They are delivery methods. They only make sense in relation to:
The capability required.
The conditions under which performance happens.
The constraints of the environment.
When modality becomes the starting point, organizations risk solving the wrong problem in the most efficient way possible.
What Learning In The Flow Of Work Actually Does Well
Learning in the flow of work is highly effective when the capability already exists. It works best for:
Recall at the moment of need.
Reinforcing known processes.
Reducing friction in execution.
Increasing consistency.
In these situations, the problem is not that people do not know what to do. The problem is that:
They cannot remember it in the moment.
They do not have easy access to it.
The process is complex enough to require support.
Here, embedding support directly into the workflow is not just useful—it is often the best option.
Where It Breaks Down
The issue arises when organizations expect flow-of-work solutions to do more than they are designed to do. They are often used in situations that require:
Judgment.
Decision-making.
Prioritization.
Adaptation under pressure.
In these cases, performance depends on capability that must exist before the moment of execution. No checklist, prompt, or embedded guide can fully compensate for a lack of underlying competence. At best, it creates dependence. At worst, it creates the illusion of capability. This becomes particularly risky in AI-enabled environments, where tools can accelerate output but cannot ensure quality or appropriateness without human judgment.
The Real Question Organizations Should Be Asking
Instead of asking: “Can we put this in the flow of work?” A better question is: “What level of capability is required for this work, and when does that capability need to exist?” From there, the decision becomes clearer:
If capability must exist before performance → it needs to be built.
If capability exists but needs reinforcement → it can be supported.
If the issue is not capability → it should be solved elsewhere.
This shifts the conversation from modality to performance.
The Risk Of Replacing One Default With Another
There is a subtle but important risk in the current trend. Organizations may believe they are evolving by moving away from courses. But if they simply replace one default with another, nothing fundamental has changed. They are still:
Selecting solutions too early.
Skipping problem definition.
Optimizing delivery rather than performance.
The tools look different. The outcomes often do not.
A More Useful Way To Think About It
Learning in the flow of work is not a strategy. It is one option within a broader set of interventions. A more effective approach is to separate three decisions:
What performance needs to improve?
What capability must exist to support that performance?
What is the least intrusive way to achieve or support that capability?
Only then does modality become relevant. And in many cases, the answer will involve a combination:
Capability built ahead of time.
Support embedded in the workflow.
Clarity in expectations and processes.
Final Thought
Learning in the flow of work is valuable. But it is not universally applicable, and it is not a replacement for capability building. When used deliberately, targeted learning in the flow of work reduces friction and improves execution. When used indiscriminately, it risks masking deeper gaps and creating false confidence.
The goal is not to choose the most modern modality. It is to choose the right intervention for the level of performance the work actually demands. That requires a different kind of discipline. One that starts with the problem—not the delivery method.
