Why don’t traditional leadership development programmes work?

Leadership development programmes don’t fail for lack of content. They fail because they are designed to take place, not to be applied. I have been doing this work for more than 20 years and I keep seeing the same scene: the company puts together a programme of two full days, flawless materials, a high NPS on the last slide and, three months later, not a trace of the expected behaviour change remains.

When I ask the HR manager what has changed in the routine of the leaders and their teams, the answer is usually “nothing noticeable”.

It isn’t ill will, and it is hardly ever a content problem. It is a design problem. The traditional model delivers information and enthusiasm, but it doesn’t create the conditions for leaders to use any of it the following Tuesday, nor for that use to show up in the team’s results.

The size of the waste

I’m an engineer by training, so I see people development as a production process. Through that lens, the numbers don’t hold up.

According to reports consolidated by global firms such as Deloitte, PwC and The Business Research Company, the worldwide corporate training market exceeds US$430 billion in 2026. The global leadership development market alone is valued at around US$100 billion in 2026, according to estimates by Future Market Insights. That is a lot of money, a lot of hours from talented people and a lot of blocked calendars.

According to Jonathan Pottiez (2024), in the study Training evaluation: manage and maximise the effectiveness of your training, 92% of business leaders don’t see the impact of learning programmes on the business, and 20% of companies don’t measure that impact at all. Despite the investment, only 13% of companies evaluate training in terms of return on investment. Yet 96% of Learning and Development (L&D) departments want to improve the way they measure learning effectiveness.

The forgetting curve explains the mechanics, and it is not an old, unverified theory. In 2015, Jaap Murre and Joeri Dros replicated Ebbinghaus’s classic experiment under controlled conditions and published it in PLOS ONE. The replication confirmed the original pattern: without any reinforcement, about a third of what was learned remains after 24 hours, a quarter after a week and just over 20% after a month. Most of the loss happens in the first few hours. The window for turning content into behaviour closes long before the follow-up we usually schedule 90 days later.

According to a Fortune survey cited by McKinsey, only 7% of CEOs believe their companies are building effective global leaders, and just 10% said their leadership development initiatives have a clear business impact. McKinsey’s own survey of more than 500 executives around the world sends a similar message: only 11% strongly agreed that their leadership development interventions achieve and sustain the desired results.

If a production line lost 90% of its output, nobody would argue for keeping the process at a board meeting. In people development, we have lived with that number for decades.

Six reasons leadership development doesn’t work

When I look at the programmes that don’t work, I see six obvious reasons, which I have been working to address with Upskill’s solutions:

1. There is no engagement, and the leader is not to blame

You invite 45 people, 20 turn up, 15 stay until the end and 5 apply something. The problem isn’t whether cameras are on or off. It is that, in live online training, nobody is doing only that. And in face-to-face sessions, with devices put away, the subject may simply be less relevant than a problem that needs solving or a deliverable with a tight deadline.

People don’t engage with what matters to the organisation; they engage with what solves a problem in their day-to-day work. When leaders haven’t chosen the topic, haven’t set the goal and can’t see where it fits into their week, they go through the motions out of obligation. But without behaviour change and results, going through the motions is expensive.

2. There is no link to business strategy

There is one question that almost never comes up in the brief: which business result changes because this leader changed their behaviour? Without that answer, the programme is built with no owner, no indicator and no defence when the budget gets tight.

The 2025 Global Leadership Development Study by Harvard Business Impact surveyed more than 1,100 L&D and HR professionals and functional leaders in 14 countries between January and March 2025. The result is uncomfortable: 4 in 10 still name alignment with business strategy as one of the main challenges of leadership development. I insist on this in my first conversation with any client, because it is what separates a programme that delivers results from just another line of expense.

3. There is no follow-up to implement what was learned

It is like a weekend cookery course: you learn a dish with three spices you can’t find in the shop near your home, you go back home and never make it again. In most programmes, the application plan boils down to a list of three commitments we ask people to write down at 5:50 pm on the last day of an immersion, for example.

That doesn’t support applying what was learned. For implementation to happen, you need to decide, before the programme starts, at which point in that leader’s routine the learning makes sense to apply, who follows up and what happens when something goes wrong on the first attempt. The leader will get stuck. It is at that point of friction that development either happens or dies.

4. It is an event, not a process

A few days, once a year. That format has hardly changed since the early 2000s. Work has changed, company structures have changed, the pace has changed, and technology has been reorganising everything, faster and faster. In most businesses, the training format remains the same. The forgetting curve shows that learning needs reinforcement points spread over time, and we keep concentrating everything into a single moment.

5. There is no customisation or personalisation

Same content, same activities, same chart for the operations manager who inherited a team in crisis and for the specialist who became a leader three weeks ago. These are two different problems, and the same class or course doesn’t properly solve either of them. Group training, whether recorded or live and in person, may address solutions to problems some leaders are facing, but many find no use in the shared content.

Customisation ensures that the shared content is aligned with the company’s current context, its culture, its moment and its strategy for delivering results. Personalisation adapts the content to each leader’s moment, focusing on what is most useful in their work routine to achieve the results the business expects.

On top of that, individual follow-up gives leaders room to reflect on the results of their default behaviour, or on the new behaviour they are putting into practice. This is different from a course with more time spent on content than on reflection, which makes it harder to understand how the knowledge can be applied successfully in each person’s reality.

6. There is no alignment with the culture and the recognition system

This is the least discussed. You spend two days teaching listening, delegation and feedback, and then send the leader back to a company that promotes people who deliver results by steamrolling their team. The programme says one thing, the reward system says another, and the leader believes the system. Development that doesn’t speak to the stated values, the promotion criteria and the recognition model doesn’t change behaviour.

What changes when learning becomes part of the routine

I don’t share these numbers to discourage those who work in HR, organisational development and L&D. I share them to change the starting point of planning.

Learning needs an action plan, not a calendar of events

And that plan doesn’t live in isolation. It is tied to business strategy, upholds and recognises the same values the company declares, and speaks to the reward system. The behaviour the programme develops needs to be the behaviour the organisation promotes, praises and pays for. When that fits together, content becomes routine and routine becomes culture, whether to strengthen it or to transform it in line with the business’s goals.

It took me a while to understand one thing: a good idea doesn’t beat routine. We think it does because we leave the training on a dopamine high, convinced that reality will behave the way we imagine. On Monday, urgency swallows everything, including the urgency created by the time the training itself took up. A good idea only survives if someone sets aside a moment for it to become action.

The product is continuity, not content

Content today is a commodity. What is rare, and usually expensive, is someone supporting the leader at the exact moment they get stuck.

That is what I pursued when designing Upskill’s model. Mentoring with a human mentor is where transformation happens: it is there that people open up, think things through together and leave with a concrete mission for the week. Between sessions, mentorIA sustains the work. It is available 24/7, knows the topics being developed and responds at the moment of the real question, not three weeks later. The human transforms. AI sustains and amplifies.

When we measure it, it shows. At an agribusiness company we work with, in a shrinking market and with leaders who needed to build resilience and self-confidence, the programme delivered a 15% increase in performance and +28% progress on goals in the very first 3-month cycle. Not because of new content, but because of diagnosis, personalisation and follow-up until the behaviour became a habit.

What you can do right now

If you lead an area or are responsible for development in your company, I suggest seven moves:

1. Before designing or hiring any programme, write down the sentence: “this programme exists so that indicator X improves”. If you can’t fill in the X, don’t hire it yet.
2. Ask whether training really is the solution. Sometimes it is a process. Sometimes it is a tool. Sometimes it is replacing the wrong person in the seat, and no programme solves that.
3. Let leaders choose their own development goal and connect that goal to their target in the organisation. Engagement starts there.
4. Design the application together with the content, within the work that already exists. Don’t create yet another meeting.
5. Build an implementation plan with reinforcement points and reviews of what has been applied.
6. Measure behaviour change, not satisfaction. An end-of-course evaluation is a fever thermometer. A low temperature doesn’t mean a healthy patient.
7. Curate time, not just content. TalentLMS’s 2026 annual L&D report names lack of time as the main obstacle to training for the third year running: 50% of learning leaders and 53% of employees say their workload leaves no room for training, even when the need is clear. And 46% of employees and 49% of HR managers acknowledge that their company treats training as time away from “real” work. As long as that is the case, no content will save the programme.

So where should a new leadership development programme start?

Leadership doesn’t happen in the auditorium. It happens on Tuesday morning, in the middle of a difficult conversation, in a decision made with incomplete information, in a conflict between two people on the team.

If development doesn’t reach that point, it isn’t development. It is an event. And we already have far too many events.

Before designing, renewing or hiring any programme, answer one question that comes before all the others: what stage of maturity is your leadership at today? Without that diagnosis, every programme is a gamble.

That is why we created the Upskill Leadership Maturity Assessment. In just a few minutes, you can map where your leadership stands. Start with your self-assessment!

→ Take the Leadership Maturity Assessment