Your Leadership Programme Is Producing Knowledge, Not Behaviour. Here Is the Design Reason.

There is a pattern that appears, with remarkable consistency, in the post-programme evaluations of corporate leadership development initiatives across India. Participants rate the content as excellent. They rate the facilitation as engaging. They describe the experience as valuable and recommend it to their peers. And then, measured six months later through direct observation, 360-degree stakeholder feedback, or simple behavioural tracking, their leadership behaviour has not changed in any significant way. The programme produced knowledge. It did not produce behaviour. And yet it received a high satisfaction score, which means it will probably be repeated in exactly the same form next year.

This is not a content quality problem. The content in most reputable leadership programmes, the frameworks, the research, the models, is genuinely good. The problem is a DESIGN ASSUMPTION, one so deeply embedded in how most L&D functions operate that it is rarely examined. The assumption is that if you give an intelligent, motivated leader access to the right knowledge and frameworks, they will translate that knowledge into changed behaviour. This assumption is wrong. Not sometimes, not in certain contexts, but structurally and predictably wrong. Albert Bandura demonstrated why in the 1970s. Donald Kirkpatrick operationalised why in his evaluation model. The knowing-doing gap, documented exhaustively by Jeffrey Pfeffer and Robert Sutton, is not a mystery. It is a design outcome.

The dominance of content delivery in leadership programme design is not accidental. It serves a number of legitimate organisational interests that have nothing to do with behaviour change. It is easier to design: a content curriculum can be built in weeks, whereas a behaviour change architecture requires deep diagnostic work, practice design, feedback infrastructure, and sustained intervention over months. It is easier to evaluate: participants can tell you whether the content was good immediately after the session, whereas measuring behaviour change requires a longer timeframe and more complex measurement tools. It is easier to justify to procurement: a programme with a recognisable content framework, a credentialed speaker, and a branded methodology is more legible in a vendor selection process than a programme whose primary value is in what happens between sessions.

There is also a psychological factor that is rarely named but consistently present. CONTENT DELIVERY FEELS PRODUCTIVE. The facilitator stands at the front of the room, shares a framework, illustrates it with examples, invites a discussion. Participants engage actively. The energy in the room is high. Everyone leaves with notes, slides, and a set of ideas they intend to apply. The session feels like something significant happened. And something did happen: knowledge transfer occurred. What did not happen is the harder, more uncomfortable, slower work of producing a new behaviour under real-world conditions. That work requires vulnerability, repetition, failure, feedback, and time. It is not the experience that generates high satisfaction scores in a post-programme survey.

L&D leaders, operating inside organisations that measure the quality of their programmes primarily through participant satisfaction ratings, are therefore structurally incentivised to design programmes that feel valuable rather than programmes that are effective. This is not a failure of individual competence or intention. It is a consequence of a measurement system that rewards the wrong outcomes. The programme that challenges participants, that sits with discomfort, that requires them to practise in front of peers and receive direct feedback on their performance, is often rated lower in the moment than the programme that inspires them with compelling content and sends them home feeling good. And yet, measured at six months, the challenging programme produces measurably more change.

Albert Bandura’s social learning theory, now more than five decades old and consistently supported by subsequent research, makes a set of claims about behaviour change that most programme designers pay lip service to and then design against. The first is that behaviour change requires not just knowledge but SELF-EFFICACY, the belief that one is capable of performing the target behaviour. Self-efficacy is not built by receiving information. It is built by experiencing small, progressive successes in performing the target behaviour. This means that a programme designed to produce more effective delegation must include opportunities for participants to actually practise delegation, to experience it going well in a controlled environment, before they are expected to perform it confidently in the real world. The delegation module in most leadership programmes does not include any practice. It includes a framework, a case study, and a discussion.

The second claim from social learning theory is that behaviour change is reinforced by OBSERVATION OF MODELS, specifically models who are similar to the learner and who are seen to succeed using the target behaviour. This is why peer cohort learning is not merely socially valuable but developmentally essential. When a participant watches a peer, someone at the same level, facing the same organisational dynamics, successfully navigate a difficult feedback conversation in a role play, the message received is not just that the behaviour is possible but that it is possible for someone like me. This is a fundamentally different message than watching a facilitator demonstrate the same behaviour with expert facility. The expert model is inspiring but not efficacious. The peer model is efficacious because the social proof is relevant.

The third, and most practically important, claim is that behaviour change requires STRUCTURED REINFORCEMENT in the context in which the behaviour is needed. This means that the learning must be transferred to the work environment and practised there, with some form of feedback, before it becomes a stable new pattern. The reason most programmes fail to produce this is that they are designed as events rather than journeys. The event ends. The participant returns to work. The work environment presents the same stimuli that have always triggered the old behaviour. Without a structural intervention that creates a new response in the old context, the old behaviour reasserts itself, not because the participant has failed to learn but because the environment has not been redesigned to support the new pattern.

Donald Kirkpatrick’s four-level training evaluation model, published in 1959 and refined repeatedly since, remains the most widely cited framework in L&D and the most systematically ignored. The four levels, for those who need the reminder, are reaction (did participants like it?), learning (did participants acquire the intended knowledge or skills?), behaviour (did participants change their on-the-job behaviour?), and results (did the behaviour change produce the intended business outcomes?). Kirkpatrick was explicit that Level 1 evaluation is the least valuable and most commonly used. He was also explicit that the absence of Level 3 measurement, behaviour change on the job, makes it impossible to know whether a programme produced anything of lasting value.

The vast majority of corporate leadership programmes in India are evaluated at Level 1 only. Some include a Level 2 assessment in the form of a knowledge check at the end of the programme. Very few conduct a Level 3 evaluation at ninety days or six months. Almost none reach Level 4 with any rigour. The implication of this is that the organisation’s understanding of what its leadership development investment is producing is based entirely on how participants felt about the experience. This is a comfortable state of affairs for programme designers and vendors, because satisfaction is manageable. It is a deeply uncomfortable state of affairs for organisations that are spending significant money on leadership development and have no mechanism for knowing whether it is working.

The argument that behaviour change is hard to measure is legitimate but not sufficient to justify the absence of any attempt to measure it. 360-degree stakeholder feedback tools exist. Behavioural observation frameworks exist. Manager assessment rubrics exist. None of these are perfect instruments, but all of them are more useful than a participant satisfaction score as an indicator of whether leadership behaviour has changed. The organisations that commit to Level 3 measurement, even imperfectly, create a fundamentally different relationship with their L&D investment. They begin to ask questions about design that organisations measuring only at Level 1 never need to ask. And in asking those questions, they create the conditions for programmes that actually change the behaviours they are designed to change.

A programme designed to produce behaviour change, rather than content absorption, makes a specific set of design choices that distinguish it from the standard model. The first and most important is that it begins with a SPECIFIC BEHAVIOUR AGENDA for each participant, not a generic content curriculum for the cohort. This requires a diagnostic phase before programme design can be completed: a 360-degree feedback process, a direct manager input, a self-assessment, or ideally all three, to identify the two or three specific behaviours that each participant most needs to develop. The programme content is then selected and sequenced in service of those specific agendas, not as a standalone curriculum that every participant receives regardless of their individual development needs.

The second design choice is PRACTICE BEFORE OUTPUT. Every concept, framework, or model introduced in the programme is immediately applied in a structured practice activity before the session moves on. The practice activity is designed at the appropriate level of difficulty, enough discomfort to stretch but not so much as to shut down learning, and followed immediately by structured feedback, ideally from both the facilitator and peers. The feedback is specific, behavioural, and forward-looking: not “that was good” or “that was difficult” but “when you did X, the impact was Y; the next time, try Z.” This cycle of practice and behavioural feedback is the primary driver of skill development in experiential learning design, and it is the element most consistently absent from content-heavy programmes.

The third choice is STRUCTURED TRANSFER to the work context. Before each participant leaves a session, they make a specific commitment about one behaviour they will practise in a specific real-world context before the next session. The commitment is as concrete as possible: not “I will delegate more” but “On Thursday, when Priya brings me the analysis to review, I will read it and ask her what she thinks before offering my own view.” This commitment is tracked at the opening of the next session, with a brief accountability conversation in a peer pair or small group. The tracking is not punitive; it is developmental. The question is not “did you do it” but “what happened when you tried it, and what did you learn?”

The responsibility for the predominance of content-over-behaviour programme design does not rest with vendors alone. L&D leaders who commission programmes without requiring evidence of behaviour change design, without insisting on Level 3 evaluation, and without building the organisational support structures that transfer enables, are complicit in producing the outcome they then complain about. This is not a criticism of individual L&D professionals, many of whom understand the problem perfectly well and are constrained by the budget, time, and measurement cultures of their organisations. It is an observation about the systemic incentives that produce suboptimal design at scale.

The most important questions an L&D leader can ask a programme vendor are not about the content: what frameworks do you use, what are your facilitators’ credentials, what is the structure of the day? The most important questions are about the design logic. What specific behaviour change outcomes is this programme designed to produce? How will we know, at six months, whether those outcomes were achieved? What is the mechanism by which content introduced in the programme gets transferred to behaviour in the work context? What structural supports does the programme include, between sessions, to sustain the transfer effort? These questions are not hostile to good vendors. They are the questions that good vendors welcome, because they create the conditions for design conversations that genuinely serve the client’s developmental goals.

Organisations that ask these questions will discover that not many programme vendors have good answers to them. The market is dominated by content delivery. Behaviour change design is rarer, harder, more expensive to produce, and less immediately legible to procurement processes that evaluate programmes by their content credentials. But it is, ultimately, the only product worth buying if what the organisation actually wants is leaders who lead differently. The knowing-doing gap is not a participant failure. It is a design failure. And design failures are correctable, given the will, the measurement rigour, and the readiness to choose discomfort over satisfaction in the service of real change.

RK

Founder & Principal Consultant, ProventusHR

RK holds the Master Certified Coach (MCC) credential from the ICF and is a Certified Practitioner of the Marshall Goldsmith Stakeholder Centred Coaching methodology. He has facilitated over 400 leadership, culture, and coaching programmes across India’s most complex organisations.