I built the Leadership Capability Stack after the same scene played out with client after client: a good leader gets sent on a course to fix a problem the course was never designed to touch, nothing changes, and the room quietly concludes the leader was the problem. They weren't. The structure underneath them was, and nobody had a way to see that until they went looking below the behaviour.
My working position is blunt: leadership performance is not developed, it's built. Those verbs point at different mental models. Developed suggests something that grows on its own if you water it with enough workshops. Built implies sequence and load: some things have to exist before other things can stand on them. Skills sit at the very top of that sequence. They matter, but they are the last piece, not the first, and most organisations start their leadership spending at exactly the wrong end.
The Leadership Capability Stack is the diagnostic I built to correct that. It has five layers: Strategic Intent at the base, then Leadership Architecture, Decision Systems, Capability Standards, and Culture Output sitting on top. Each layer depends entirely on the integrity of the one beneath it, and once you start reading organisations through that lens, the pattern behind most failed transformation programmes becomes hard to unsee.
The Five Layers of the Leadership Capability Stack
- Strategic Intent: The foundation layer. What the organisation is actually trying to achieve, specific enough to guide a daily decision, stable enough to create predictability, and adaptive enough to move when the market does. Not a value painted on a lobby wall.
- Leadership Architecture: The structural blueprint built on top of intent: how leadership roles, levels, and accountability are designed, and who owns what once the org chart stops being decorative.
- Decision Systems: How authority, decision rights, and escalation paths are wired so strategic intent converts into action, instead of stalling in a meeting nobody has the standing to end.
- Capability Standards: What good leadership looks like at each level, only meaningful once it is wired into the structure beneath it. Disconnected from that structure, it is a generic competency list nobody applies under pressure.
- Culture Output: The top of the stack. Behaviour and culture are what the four layers beneath produce. You engineer it indirectly by fixing the machine underneath, not by pulling on culture itself.
What is the Leadership Capability Stack?
The Leadership Capability Stack is a five-layer diagnostic model for working out why leadership performance varies so widely inside a single organisation, often between two teams reporting into the same executive with the same training budget behind them. I borrowed the layered metaphor deliberately from how software runs: an application sits on a technology stack of interdependent layers, and a failure low in that stack takes the whole application down, no matter how well the top layer was coded. Leadership behaves the same way. A team can have excellent individual leaders and still perform inconsistently, because the fracture sits underneath them, not in them.
The sequence runs ground up:
- Strategic Intent: what the organisation is trying to achieve, clear enough to guide daily decisions without a follow-up meeting.
- Leadership Architecture: how leadership roles are designed and where accountability sits, level by level.
- Decision Systems: how authority and escalation are structured so intent turns into action instead of stalling in ambiguity.
- Capability Standards: what good leadership looks like at each level, useful only when it connects to the structure above it.
- Culture Output: the behaviour a healthy or unhealthy stack produces. The result you read last, never the lever you should pull first.
Culture sits at the top on purpose. It is a readout, not a control panel. I have watched plenty of culture complaints resolve on their own once Strategic Intent and Decision Systems got fixed and nobody touched culture directly at all, and I have never once watched a culture initiative succeed while the layers underneath it stayed broken.
Why does leadership development fail without a structural diagnosis first?
Organisations spend enormous sums trying to fix this at the wrong layer. Harvard Business Review researchers Michael Beer, Magnus Finnström, and Derek Schrader put a number on it: American organisations spent roughly $160 billion domestically and close to $356 billion globally on training in a single year, and their verdict was direct, the learning rarely lifts organisational performance because people revert to their old habits within days of the workshop ending (Beer, Finnström & Schrader, Harvard Business Review, 2016). That is what happens when you polish the Capability Standards layer while Strategic Intent or Decision Systems underneath it stays cracked. A training-quality problem it is not.
The pattern shows up in the numbers on leadership quality itself, not just training spend. DDI's Global Leadership Forecast found that only 40 percent of leaders rated their own organisation's leadership as high quality, a 17-point fall in just two years, the steepest drop the research has recorded in a decade (DDI, Global Leadership Forecast 2023). Two years is too short a window for a whole generation of leaders to have gotten measurably worse at their jobs. What actually happened underneath them: the structures they operate inside grew more complex faster than anyone rebuilt the architecture to hold that complexity.
Send a capable leader on a course to fix a problem actually caused by unclear decision rights, and nothing changes, because the skill was never the missing ingredient. The complaints return after training when the intervention misses the layer that fractured. You can retrain the same cohort three years running and hear the same complaints in next year's engagement survey, because the survey measures an output and nobody has gone near the inputs producing it.
Instead of asking, "how do we develop better leaders," the Stack forces a sharper question first: have we built the structural foundation that lets leadership excellence exist at all. That reframe carries the whole value of the diagnostic. It points down toward the cause instead of up at the symptom, and it is uncomfortable, because the honest answer is usually a conversation with the executive team about whether Strategic Intent is actually clear, and most executive teams would rather sign off on another training budget than have that conversation.
What is the foundational layer of the Stack?
Strategic Intent is the base layer, and the one most organisations assume is solid right up until you test it. It answers a question that sounds almost too simple to matter: what is this organisation actually trying to achieve?
This is not a mission statement laminated in reception. Strategic Intent is crystallised clarity about where the organisation is heading, why that destination is worth the cost of getting there, and what success looks like on arrival. Specific enough to settle a daily decision without a follow-up meeting. Stable enough to create predictability across quarters. Adaptive enough to move when the market moves, rather than snapping the moment it is tested.
Without it, every layer above becomes unstable. Leaders cannot make aligned calls without knowing the priorities behind them. They cannot allocate scarce time and budget without agreement on which capabilities matter most this year. The result is what I call leadership drift: capable people working hard toward individually sensible, collectively divergent goals, each one convinced they are executing the plan. The strongest leadership team on the planet cannot out-execute this kind of failure, because no one can coordinate around a destination that was never agreed in the first place.
Strategic Intent is the gravitational centre that keeps every leadership decision in orbit around a shared destination. Most leadership teams write it once, frame it, and never test it again.
How does Leadership Architecture sit on top of intent?
Leadership Architecture is the structural blueprint built on Strategic Intent, the design of who leads at each level and what they are actually accountable for across the organisation. Solid intent gives this layer something real to align to. Without it, the architecture is just an org chart drawn to look tidy in a slide deck.
This layer asks the structural questions nobody wants to schedule a meeting about. How many leadership levels genuinely exist versus how many exist on paper? What unique value does each level actually add? Where do two people quietly own the same decision, and where does nobody own it at all?
Most organisations inherit their leadership architecture rather than design it, and that inheritance is the trap. Layers accumulate across a decade of growth, acquisition, and reorganisation, each one adding complexity without adding clarity. The Stack makes that accumulated mess visible so it can be redesigned on purpose instead of endured by habit. The expensive dysfunction often sits here: a decision with three quiet co-owners, or none, while everyone assumes somebody else has it.
Why does Decision Systems break more leadership programmes than any other layer?
Decision Systems is where strategic intent either becomes action or stalls out in a meeting nobody has the authority to close. This layer governs who can decide what, how far a decision can travel before it needs sign-off, and what happens when two capable people disagree about the right call.
Bain & Company's research on decision effectiveness makes the same point I make with clients constantly: fast, well-owned decisions and clean execution are what actually separate the strongest financial performers from everyone else, not smarter people in the room (Bain & Company, Measuring Decision Effectiveness). Most organisations do not have a decision-making problem in the sense of poor judgement. What they have is a decision-ownership problem: too many people with a vote, too few with the authority to close the conversation and move.
You can see a broken Decision Systems layer without opening a single dashboard. Watch how long it takes a routine call to get made, and count how many people have to weigh in before it does. If that number keeps climbing quarter over quarter, the architecture above it may look fine on paper while quietly rotting underneath.
How do you use the Stack to diagnose a real leadership problem?
Diagnosis means matching the visible symptom to its likely layer, then working downward until you find the actual fracture. The symptom on the surface is almost never where the damage sits.
- Decision paralysis or chaos usually points to a Decision Systems failure: unclear authority and escalation paths.
- Role confusion and duplicated ownership point to a Leadership Architecture problem.
- Initiatives pulling in different directions mean Strategic Intent is unclear or never cascaded past the top team.
- Inconsistent leadership behaviour means Capability Standards exist without being wired to the structure above them.
- Cultural dysfunction is almost never a behaviour problem to solve directly; it is the visible output of a fractured foundational layer.
Then apply the rule that saves the most wasted budget: work the stack from the bottom up. Investing in Capability Standards while Leadership Architecture is still broken just builds skills the structure cannot use. Fix the foundation first, though quick wins inside Decision Systems can build momentum while the deeper architectural work is underway. Diagnosis before design. Design before development, in that order, every time.
The Stack Diagnostic: four tests to find the layer that broke
The Stack Diagnostic
- The retrain test: Has this exact team been trained on this exact issue before, more than once? A repeat complaint after a repeat course is proof the fracture sits below Capability Standards, not inside it.
- The three-owner test: Ask three people who owns a specific recurring decision. If you get three different names, or none, you have a Leadership Architecture problem, not a personality clash.
- The escalation test: Count how many decisions in the past month travelled two or more levels above where the relevant information actually lived before anyone signed off. Each one is a Decision Systems failure, dressed up as caution.
- The workshop-relapse test: Did a culture score move for one quarter after an initiative, then quietly slide back? That is a Culture Output symptom being treated at the top of the stack instead of at the foundation producing it.
None of these tests are soft, and I use them precisely because a leader who has simply learned the right vocabulary cannot talk their way past them. Anyone can say the words "we have clear decision rights." Far fewer can name, on the spot, who owns the decision that came up in yesterday's leadership meeting.
How is the Stack different from a competency framework or a general leadership model?
A competency framework operates only at the Capability Standards layer, describing what good leadership looks like without ever asking whether the structure beneath it can hold that behaviour up. That is why so many well-written competency frameworks sit in a drawer: they describe a destination without diagnosing what is blocking the road there. The Stack is the missing layer underneath a competency framework, the part that decides whether it has anything real to attach to.
It also sits differently to a values or culture initiative, which usually starts at the top of the stack and tries to work down through behaviour change alone. I have watched that approach produce a genuinely moving all-hands meeting and zero change in how decisions actually get made three months later, because nobody touched Strategic Intent or Decision Systems, the two layers actually generating the culture everyone was trying to talk their way out of.
If you want the deeper mechanics of how the Leadership Architecture layer gets designed on purpose rather than inherited by accident, or how the Stack holds up once an organisation is scaling fast enough to outrun last year's structure, those are their own bodies of work built on top of this one: see how leadership capability scales through high-growth stages, and how the same five layers apply to building capability in people much earlier in their leadership path.
Questions senior leaders need to answer
If you remember one thing, remember this: leadership problems are architecture problems wearing a behaviour costume. Every time I have been called in to fix "a culture issue" or "a leadership gap," the real work has started two or three layers below wherever the client first pointed. That is what happens every time you trace a symptom back to its structural source instead of medicating it where it happens to show up, and it has nothing to do with a coaching trick.
I will go further than most people in this field are willing to. Coaching an individual leader without first checking the architecture around them is, more often than not, a waste of a good leader's time and the organisation's money. Coaching itself works, when the structure can hold what it produces. Pour a well-coached leader back into a system with unclear decision rights and no shared strategic intent, and you have made one person briefly more self-aware inside a machine that was never built to use that self-awareness, and probably burned them out a little faster in the process.
My sequence starts with the stack, then the intervention, before any development spend is approved. Organisations that skip straight to development get a temporary lift and a familiar relapse eighteen months later. Organisations willing to do the diagnostic work first get something that holds, because they are finally building on ground that can bear the weight.
So when someone asks what the Leadership Capability Stack is, my answer is not the five-layer diagram. The diagram is just the tool. My answer is the discipline underneath it: stop asking people to be better, and start asking whether you have built a structure in which better is even possible. Most organisations have not. Call that an architecture failure, not a character one, and architecture is fixable once you are willing to look below the floor you are currently standing on.
Three places to go next if you want to apply this directly: the full breakdown of the Architecture Accelerator programme, where I run this diagnostic with executive teams end to end, CapabilityAI, built to help a leadership team apply Stack-level thinking to a live decision between sessions, and, once you have found the layer that broke, The Leadership Operating System Playbook, which covers how to actually run the fix day to day once you have diagnosed it.