The essentials at a glance
- Good leadership development is a system, not an event. It should move people through diagnosis, practice, feedback, and review.
- Leadership and management overlap, but they are not the same. A useful framework develops both execution and influence.
- Experience does most of the heavy lifting. Formal learning matters, but stretch work and developmental relationships create deeper change.
- Role level matters. First-line managers, middle managers, and senior leaders need different capabilities and different support.
- Measurement should go beyond attendance. Behaviour change, team outcomes, retention, and readiness for promotion tell a better story.
What a good model needs to do
When I look at a leadership framework, I ask a simple question: does it change how people behave when work gets messy? A useful model should do three things at once. It should identify the capabilities that matter for the organisation, give people a realistic way to practise them, and create evidence that those capabilities are showing up in day-to-day work.
That is also where leadership and management split slightly. Management keeps work organised, resources allocated, and deadlines under control. Leadership sets direction, creates alignment, and helps people move through uncertainty. Strong development plans treat both as essential, because a manager who cannot prioritise or delegate will struggle, but so will someone who can run a process without influencing others or making sound judgement calls.In practice, I prefer models that are built around observable behaviours rather than vague traits. “Strategic thinking” is too broad on its own. “Can translate a business goal into a team plan” is better. “Communicates well” is not enough either. “Holds difficult conversations early, before problems grow” is the sort of behaviour that can actually be coached. Once that distinction is clear, the process becomes much easier to design.
The stages I would build into the process
The best models I have seen follow a fairly simple sequence. The order matters more than most teams expect, because jumping straight to training usually creates generic learning instead of targeted growth.
| Stage | What it should include | What success looks like |
|---|---|---|
| Diagnose | 360 feedback, manager input, self-assessment, performance data, and succession needs | A clear view of the exact behaviour gaps to close |
| Design | Role-specific goals, stretch assignments, coaching, and targeted learning content | A development plan linked to a real business need |
| Learn | Short learning modules, simulations, mentoring, observation, and peer discussion | People understand both the concept and the practical application |
| Apply | Live projects, delegated responsibility, difficult conversations, and decision-making under pressure | New behaviour shows up in real work, not only in workshops |
| Review | Feedback loops, reflection, and outcome tracking after 3 to 12 months | Evidence of behaviour change, not just satisfaction scores |
If one of those stages is missing, the model weakens quickly. Without diagnosis, everyone gets the same content. Without application, the learning stays abstract. Without review, managers cannot tell whether the effort was worth it. That sequence matters because leadership grows through repetition, not through a single memorable session, which is where the learning mix comes in.
Why the 70-20-10 idea still matters
I still find CCL’s 70-20-10 framework useful because it forces organisations to stop overvaluing classroom training. The numbers are not meant to be treated like accounting rules; they are a reminder that leaders learn from three sources: challenging experience, developmental relationships, and formal coursework.
| Learning source | What it contributes | What goes wrong if it dominates |
|---|---|---|
| Challenging assignments | Judgement, resilience, problem-solving, and context-specific learning | People are thrown in at the deep end without support |
| Developmental relationships | Perspective, feedback, sponsorship, and correction in real time | Growth becomes dependent on one helpful person |
| Formal learning | Shared language, frameworks, tools, and a safe place to rehearse ideas | The learning stays theoretical and fades quickly |
The practical lesson is straightforward. I would never build a leadership programme around content alone, because content without application rarely changes behaviour. At the same time, I would not rely only on stretch assignments, because challenge without feedback can turn into confusion. The strongest version blends all three so the training acts as an amplifier, not the whole engine.
That approach also helps different audiences. A new manager may need more formal learning than a seasoned director, but both still need real assignments and honest feedback. The model becomes far more useful once you adapt it to the realities of role, sector, and workload.
How I would adapt it for UK organisations
In the UK, I would pay close attention to the middle of the organisation. That is often where strategy is either translated well or lost in the noise. Front-line managers need confidence with delegation, feedback, wellbeing conversations, and day-to-day accountability. Middle managers need cross-functional influence, prioritisation, and the ability to lead change without losing delivery. Senior leaders need more work on judgement, culture, and succession than on basic people management.
Hybrid work adds another layer. When teams are not co-located, leadership is tested by how clearly priorities are set, how consistently communication happens, and how early problems are surfaced. In that setting, a development model that depends entirely on observation in the office is too weak. I would build in remote coaching, peer learning, and structured check-ins so the learning continues between formal touchpoints.
| Level | Primary focus | Best development actions |
|---|---|---|
| First-line managers | Delegation, feedback, team rhythm, and confidence | Shadowing, live coaching, and short practice cycles |
| Middle managers | Influence across teams, prioritisation, and change delivery | Stretch projects, peer groups, and mentor support |
| Senior leaders | Strategy, culture, succession, and ethical judgement | Executive coaching, scenario planning, and board-level exposure |
| Emerging talent | Self-leadership, initiative, and influence without authority | Rotational work, project ownership, and mentoring |
I also would not create a programme that only serves the top layer. In many organisations, that is where budgets go, but the real bottleneck is often a weak manager population one or two levels down. If those managers cannot coach, decide, and communicate clearly, the whole system slows down. That is why design choices matter as much as content, and why the most common mistakes are worth naming directly.
The mistakes that quietly break the model
Most poor results come from avoidable errors rather than bad intent. The first is treating leadership development as a one-off event. A workshop may create energy, but without follow-up it rarely changes behaviour. The second is promoting high performers and assuming they will simply “grow into” leadership. Performance and leadership potential overlap, but they are not the same skill set.
A third mistake is overloading people with content and underloading them with practice. I see this often: slides, frameworks, and reading lists, but very little rehearsal of actual leadership moments. Another weak pattern is designing one generic programme for everyone. A newly promoted supervisor does not need the same intervention as a senior leader managing succession, risk, and culture. Leadership development also fails when line managers are left out of the process, because they are usually the people who reinforce or block new behaviour.
The evidence matters here. A CIPD review found that leadership training has a moderate positive effect, but the impact varies by participant and setting, and people who already use strong leadership techniques often gain less from extra training. In other words, more training is not automatically better. The right training, reinforced in the right way, is what creates progress. Finally, I would watch for vanity metrics. Attendance, completion, and satisfaction are useful as hygiene checks, but they do not tell you whether someone is actually leading better. That is why measurement has to sit inside the model from the start, not at the end.How I would measure whether it is working
I would measure leadership development on three levels: behaviour, team outcomes, and business impact. Behaviour shows whether people are using the skills. Team outcomes show whether those skills matter in the real world. Business impact shows whether the organisation is getting a return on the investment.
| Metric type | Examples | Why it matters |
|---|---|---|
| Behaviour change | 360 feedback, manager observation, coaching notes, and self-reflection | Tells you whether the person is leading differently |
| Team outcomes | Retention, engagement, absenteeism, team performance, and delivery consistency | Shows whether the new behaviour is helping people around them |
| Pipeline strength | Promotion readiness, internal fill rate, and succession coverage | Reveals whether the organisation is building future leaders |
| Business impact | Project delivery, customer outcomes, fewer escalations, and productivity gains | Connects the model to operational results |
If I need a simple ROI view, I use a basic formula: (benefits minus cost) divided by cost, multiplied by 100. That is not the whole story, because some benefits are harder to price, such as better collaboration, inclusion, or decision quality. Still, it is a useful discipline. A model that cannot show any movement after several months is probably too generic, too shallow, or too disconnected from actual work.
For most roles, I would look for changes over at least 3 to 6 months, and for more senior positions, 6 to 12 months is often more realistic. Leadership rarely changes on the timetable of a training calendar. It changes when people are asked to do difficult things, receive feedback, and then do them again more skilfully.
What I would prioritise first in a new programme
If I were building a new leadership programme from scratch, I would keep the first version deliberately small. I would define four or five behaviours that matter most, link them to real business outcomes, and give each participant one stretch assignment that forces those behaviours into daily use. That is usually more effective than launching a broad curriculum full of well-meaning topics.
- Start with the role, not the theory. The best development plans are tied to specific responsibilities and decisions.
- Use managers as coaches. If direct managers do not reinforce the learning, the model loses momentum.
- Mix challenge with support. Stretch work without feedback is risky; feedback without challenge is inert.
- Track a few meaningful measures. Behaviour change, retention, and readiness for promotion are enough to begin with.
That is the version of leadership development I trust most: clear expectations, real work, consistent feedback, and enough structure to make progress visible. When those pieces are in place, the model stops being a training calendar and starts becoming a system that changes how people lead.
