Your best manager was probably not your best manager. Your best manager was the one who handed you something you weren't ready for and then stayed close enough to catch you if it fell apart.
That is the whole mechanism behind leadership development through stretch assignments and projects. Not a course. Not a personality test. A task that is slightly too big, with a deadline and a person who will notice if you fail. Everything else — the framework, the coaching, the assessment — is scaffolding around that one uncomfortable moment.
I'll be blunt about where I stand: most leadership programmes fail because they're safe. People sit in rooms, they nod at slides about emotional intelligence, and they return to the exact same job with the exact same scope. Nothing is at risk, so nothing changes. The managers I've watched grow fastest were the ones pushed into work that genuinely scared them.
Key Takeaways
- A stretch assignment is work deliberately one level beyond current capability — with real consequences attached.
- The 70-20-10 rule puts roughly 70% of development in on-the-job experience, not in classrooms.
- Assignments fail when scope is vague, learning goals are unspoken, or the feedback loop is missing.
- Stretch work generates observable performance data, which makes succession decisions far less political.
- The 30-60-90 rule gives you a rhythm: learn, contribute, then own.
Why stretch assignments beat training courses for building leaders
Ask a room of managers where they actually learned to handle conflict, budget pressure, or a failing team. Almost nobody names a workshop. They name a project that went sideways, a client who nearly walked, a merger nobody understood.
The reason is mechanical, not mystical. Sitting in a course, you practise judgement with no cost attached to being wrong. On a stretch project, your decision about which supplier to cut or which hire to approve lands on real people. That stakes level is what makes the learning stick — and it's exactly what a classroom can't simulate.
The capability gap you should aim for
Too small a gap and the person coasts. Too large and they drown, which teaches them only that they were set up. My rule of thumb after watching plenty of these go wrong: the assignment should feel roughly 70% achievable with the skills they already have. The remaining 30% is where growth happens, and it has to be reachable through effort, mentoring, or a bit of hard thinking.
One example I keep coming back to. A capable individual contributor on my team — sharp analyst, terrified of speaking up — was given the job of presenting our quarterly findings to the executive committee. Not "help prepare." Present it. She was visibly miserable for two weeks. On the day she was decent. The quarter after, she was good. By the third quarter, she was the person other teams asked for. No course on presentation skills would have delivered that, because no course would have put her name on the agenda.
What goes wrong when nobody designs it
The failure mode I see most often is the assignment that isn't really an assignment. Someone gets handed extra work with no framing, no authority to make decisions, and no one checking in. That's not development — that's a covert workload dump, and the person on the receiving end knows the difference immediately.
Second mistake, and I made this one myself: forgetting to say out loud that this was meant to be a stretch. I assumed it was obvious. It wasn't. The person thought they were being tested for a promotion they might fail. That small omission cost me a month of good faith before we talked it through.
What are examples of stretch assignments?
Concrete, named situations work better than abstract principles here. These are the patterns I've either used, been given, or watched work — pick the one that matches the gap you're trying to close.
- Leading a cross-functional launch where the person has zero formal authority over half the team — tests influence without hierarchy.
- Taking over a project that's already failing, with an explicit mandate to either fix it or kill it cleanly.
- Representing the department at a senior-level meeting. Short, high exposure, low duration.
- Managing a budget for the first time, including the uncomfortable part where someone has to say no to a colleague's request.
- Onboarding and mentoring a new joiner through their first three months — quiet, unglamorous, brutally revealing about how well someone actually understands the work.
- Running a difficult client relationship solo, with a clear escalation route if it truly derails.
- Standing in for a manager during extended leave, which is a genuine stretch because it doesn't come with the title.
Notice what these have in common: each one puts a decision in the person's hands. If they can't make a call that matters, it isn't a stretch assignment.
What is the 70-20-10 rule for leadership development?
The 70-20-10 rule is a widely used heuristic for how people actually develop at work: roughly 70% from challenging on-the-job experience, 20% from relationships such as mentoring and coaching, and 10% from formal courses and reading. It's a rule of thumb rather than a precise measurement, and treating the numbers as exact is a mistake. The proportions are directional.
Where the model earns its keep is in budgeting attention, not money. Most organisations do the reverse: they spend the bulk of their effort on the 10% because courses are easy to buy and easy to log. The 70% — who gets put on what — is messy, political, and requires a manager to make a judgement call about a person's readiness. That's why it gets neglected.
One caveat worth stating plainly: the 70% doesn't work without the 20%. Experience alone, unexamined, just produces habits. A mentor or manager who asks "what would you do differently next time?" turns a busy quarter into actual learning.
What are the 5 C's of leadership development?
Different frameworks use different words, but the 5 C's most commonly cited in leadership development are competence, character, commitment, communication, and confidence. Treat that list as one version among several — you'll see it vary between sources — but the underlying idea holds: leadership capability is not a single skill, and a stretch assignment is useful precisely because it touches several of these at once.
Competence is the technical and judgement skill. Character is how someone behaves when the pressure rises and nobody's watching. Commitment shows up when the project gets boring in week nine. Communication is whether they can explain a decision to people who disagree. Confidence is the quieter one — not bravado, but the willingness to make a call and own it.
Here's why this matters practically. If you design an assignment purely to test competence, you learn nothing about how that person treats people under stress. A well-chosen stretch project should create at least one moment where character and communication get tested, because that's what you're actually assessing before you promote someone.
What is the 30-60-90 rule in leadership?
The 30-60-90 rule in leadership is a staged onboarding rhythm for someone stepping into a new role or a big stretch project: the first 30 days are for learning, days 30-60 for contributing with support, and days 60-90 for taking ownership with real accountability. It stops people from either moving too fast and breaking things, or hiding for a quarter and calling it "listening."
I use it as a scaffold for stretch work because it sets expectations on both sides. In the first thirty days, the person's job is to ask the obvious questions and map the terrain — and their manager's job is to not punish them for not having answers yet. By day sixty, they should be delivering something visible. By day ninety, they should be the person making the call.
The value is in the explicit permission to be a beginner for a defined period. Without it, ambitious people skip straight to acting confident in week one, and then nobody can tell whether they actually understand the situation or are bluffing.
| Phase | Primary focus | What the manager should do |
|---|---|---|
| Days 1-30 | Learning: context, people, constraints | Open doors, answer questions, protect the learning time |
| Days 30-60 | Contributing on a defined piece of work | Give feedback quickly and specifically |
| Days 60-90 | Owning outcomes and decisions | Step back; only intervene on genuine risk |
How to make stretch work actually develop people
Three things need to be true before you hand someone a project that's beyond them. Get one wrong and you're not developing a leader, you're generating a resignation.
- Name the learning goal out loud. "This is to build your capability in X" removes the ambiguity that turns a stretch into a threat.
- Define the scope and the decision rights. What can this person decide alone, what needs a check-in, what's off limits.
- Schedule the feedback loop before you start. A 20-minute review every two weeks beats a heroic post-mortem at the end.
- Have an exit ramp. If it's genuinely going badly, there must be a way to stop it without it reading as a career failure. Sounds soft; prevents real damage.
And be honest about the awkward part. Stretch assignments generate observable performance data — who coped, who folded, who quietly fixed something nobody asked them to fix. That data makes succession conversations far less dependent on who's most visible in meetings. Slightly uncomfortable for the people who rely on visibility. Better for everyone else.
A final thought
The uncomfortable question worth sitting with: when did you last give someone work that made them nervous, and then stayed close enough to help them through it?
Most managers, if they're honest, can't remember. Not because they don't care, but because handing out stretch work means accepting that something might go wrong on your watch. That's the price. The alternative is a team that's perfectly competent at exactly what it does today, forever.
