Why Training Programs Do Not Guarantee Behavior Change
The Training-as-Solution Default
Training is the most common organizational response to behavior change challenges. When people are not doing something, the default assumption is that they do not know how. So organizations build courses, run workshops, create e-learning modules, and deploy coaching programs. When behavior still does not change, they build more training or blame the quality of the existing training.
What Training Actually Solves
Training builds Capability: knowledge and skill. When the barrier to behavior change is genuinely a capability gap (people do not know how to do the new thing or cannot do it fluently under real conditions), training is exactly the right intervention. The problem is that capability is only one of three COM-B domains. When organizations use training to address Opportunity barriers (the environment does not support the behavior) or Motivation barriers (the behavior does not feel worth doing), training consumes resources without moving the behavioral needle.
Research on training transfer consistently shows that the work environment has more influence on whether trained behaviors persist than the training content itself. If the learner returns to an environment where the old behavior is still the default, where peers are not performing the new behavior, and where the new behavior takes more effort, the training will not transfer. This is not a training quality problem. It is a system design problem.
The Diagnostic Alternative
The SHIFT framework uses COM-B to diagnose the barrier before selecting the intervention. If diagnosis confirms a capability barrier, training is the right investment. But the training should be designed with behavioral transfer in mind: implementation intentions, graduated difficulty, peer teaching, and environmental prompts that support application in the real work context.
If diagnosis reveals an opportunity barrier (the system makes the old behavior easier) or a motivation barrier (peers are not doing it, or the behavior feels risky), the intervention is not training. It might be environmental redesign, social proof strategies, friction reduction, or coalition building. These require different investments than training, and they address different root causes.
The practical rule: diagnose before you train. If the answer to 'why is this behavior not happening?' is 'they do not know how,' train. If the answer is anything else, invest in the intervention that matches the actual barrier.
Training is the right intervention when capability is the confirmed barrier. When training fails to produce behavior change, the barrier was probably not capability. Diagnose with COM-B before defaulting to more training.
