In medicine, prescribing treatment before making a diagnosis would be considered poor professional practice. The physician begins with questions, observations, and evidence. Learning and Development should operate on the same principle. Yet many learning projects begin with the proposed treatment already decided.
A typical request might sound like “We need customer service training” or “We need to retrain everyone before the next audit.” These requests may seem like learning requirements, but they are often business problems expressed through the most familiar solution available to the person making the request.
The problem is not that managers ask for training, but when L&D accepts the requested solution without testing the assumptions behind it. The discussion quickly moves to learning objectives, course length, delivery method, budget, and launch date, without ever making a diagnosis.
Training can improve performance when people lack relevant knowledge or skill, and when the workplace allows them to use what they learn. However, the environment surrounding the learner materially affects whether training appears in workplace behavior. A meta-analysis examining work-environment support found that peer, supervisor, and organizational support all had moderate, positive relationships with the transfer and continued use of trained knowledge and skills.
Manager support is also more complex than asking supervisors to “encourage participation.” A systematic review of 99 studies identified 24 distinct supervisor behaviors and attitudes that can support training transfer. Motivation matters as well, as a meta-analysis covering 148 studies, 197 effect sizes, and 31,718 participants examined the relationship between motivation and the transfer of professional training.
A well-designed course can still fail when managers do not reinforce it, employees lack opportunities to apply it, workflows discourage it, or incentives reward something else. That’s why the training request cannot be the starting point. A training request is an untested hypothesis, and L&D needs a repeatable method for investigating the problem behind it.
Related: Key Takeaways from ISTELive and ASCD 2026 Conference
The Performance Diagnosis Ladder is a five-stage diagnostic sequence that tests different assumptions. The order matters, and moving directly to the fifth stage usually means the solution is being selected before the problem is understood. The stages include: identifying the business outcome to improve, describing the observable performance required, gathering evidence of the performance gap, examining the environment, and determining the role of learning.
For example, if a business unit requests refresher training because employees are entering incorrect information into a customer-management system, the diagnosis might reveal that employees can explain the correct procedure and complete it accurately in a controlled demonstration. The capability exists, but the errors occur during high-volume periods due to a lack of clear explanation of what information belongs in a newly mandatory field.
When L&D diagnoses before designing, its conversations with stakeholders change. Instead of asking about course modules or launch dates, it asks about the evidence supporting the proposed solution and what must improve. This does not make L&D less responsive, but more useful. The strongest minority objection is that stakeholders may have a fixed deadline or compliance requirement, but diagnosis does not always require a large research project.
A 30-minute conversation, 2 employee interviews, a review of error data, and 1 workplace observation can be enough to test the central assumption. The depth of analysis should reflect the cost and risk of being wrong. Before L&D commits to development, it should be able to complete a one-page diagnosis containing an explicit performance hypothesis that can be tested.
This approach creates a new operational standard for accepting training work and provides a credibility test for L&D. By diagnosing performance issues before prescribing learning solutions, L&D can ensure that its interventions are effective and efficient, and that they address the root causes of the problems rather than just the symptoms.
They can ensure interventions are effective.
