Diagnose Before Executive Coaching: Check the System Before You Develop the Leader

In my view you should diagnose before executive coaching, because many problems that look like a leader's weakness are produced by the system around the leader: unclear decision rights, incentives that reward the wrong behaviour, a missing development pathway, a meeting rhythm that punishes candour. Coaching a leader inside a faulty system usually makes them better at coping with it, not better at changing it.

After 25 years advising CEOs and executive teams, I find the most expensive coaching decisions are the ones made before anyone has asked what the problem actually is. A leader is slow to decide, so the business buys coaching. A leader avoids conflict, so the business buys coaching. Sometimes that is right. Often the leader is slow because three people hold overlapping authority, or avoids conflict because the last person who raised an issue was marginalised. The coaching then addresses a symptom, the system stays as it was, and the next leader in the seat shows the same pattern.

Why diagnose before executive coaching?

There are three reasons, and the research supports each of them in part. The first is that the field cannot tell you in advance which design will work. Cannon-Bowers and colleagues, in a 2023 meta-analysis in Frontiers in Psychology, found that workplace coaching is effective in achieving positive organisational outcomes. They also describe the field as relatively immature, because researchers do not yet understand exactly what works, which approaches are most effective, or how long coaching needs to take. If the evidence cannot choose the approach, the buyer has to, and a diagnosis is how a buyer chooses.

That is the second reason. Learning that does not transfer to the workplace is a design problem as much as a participant problem, and Geerts groups the factors that decide whether learning transfers into programme design, delivery, personnel, evaluation and direct application, plus organisational culture and support. Several of those sit outside the leader. The third reason is cost of error. If you coach the wrong thing, you have spent the budget and, worse, told a capable leader that the problem is theirs when it was not.

Gallup offers a related data point. It estimates that managers account for at least 70 per cent of the variance in employee engagement scores across business units. I read that as a reminder of how much a leader's behaviour matters to the people around them, and therefore how much it is worth getting the diagnosis right before investing in changing it. It does not say that the leader is always the cause.

Five tests that separate a leader problem from a system problem

1. Does the pattern follow the person or the seat?: If the previous holder of the role and the person before them showed the same pattern, look at the role, its authority and its incentives. If the pattern appeared only with this leader, look at the person.
2. Do several capable people behave the same way?: One leader who avoids conflict is a personal pattern. Five leaders who avoid conflict are responding to something in the environment, such as what happened to the last person who disagreed in public.
3. What does the system reward?: List what gets promoted, praised and paid. If speed is rewarded and accuracy is not, expect speed. Behaviour that frustrates you is often behaviour the organisation has trained.
4. Is decision ownership clear?: Ask three people who owns a recent decision. If you receive three names, or none, the leader is working in an unclear structure. Coaching will not clarify it.
5. Does the leader know what good looks like at the next level?: If there is no defined pathway, a capable leader is guessing at what is expected. Coaching can help them reflect, but the missing standard is the real gap.

These tests do not settle the question on their own, and some problems are both personal and structural. They do change the first move. When the tests point to the system, I would redesign the decision rights, incentives or pathway first, and use coaching to help leaders adopt the new design. My article on when executive coaching is the wrong intervention develops this, and fixing uneven leadership quality covers the case where the symptom appears across many leaders.

Consider how the two kinds of error differ. If you diagnose a system problem as a leader problem, you coach a capable person for six months, the structure stays as it was, and either the leader leaves or the pattern returns. If you diagnose a leader problem as a system problem, you spend a quarter redesigning decision rights while the real issue goes unaddressed. The first error is more common, because coaching is easier to buy than a redesign, and because it is socially simpler to name a person than a structure. A short, honest diagnosis protects you from both errors, and it also gives you something to show the leader: here is what we found, here is why we chose this, and here is what we expect to change.

The diagnosis also changes how the leader receives the coaching. A leader told that they have a problem and are being sent for coaching reads it as a verdict. A leader who has seen the evidence, including the parts that point at the organisation, tends to approach the coaching as a shared piece of work. That difference in framing shows up in candour from the first session, and candour is what the coach has to work with.

Applying Leadership Capability Architecture to the question

Leadership Capability Architecture is the lens I use to sort the two. It treats a leader's performance as the product of a system, and my free Leadership Diagnostic scores five structural pillars: the Alignment System, the Leadership Identity System, the Execution Cadence System, Capability Pathways and the Cultural Reinforcement System. Each pillar points to a different first move.

• Alignment System: if priorities and decision rights are unclear, clarify them. Coaching will not.
• Leadership Identity System: if leaders struggle with how they see their role, such as a new CEO who still thinks like a functional head, coaching is a good fit.
• Execution Cadence System: if decisions do not carry through because meetings and reviews do not link up, redesign the cadence.
• Capability Pathways: if there is no defined route from one level to the next, build it and then use coaching and experience to move leaders along it.
• Cultural Reinforcement System: if the organisation punishes the behaviour you want, change what is rewarded before asking leaders to behave differently.

The sequence in my 5D framework is Discover, Desire, Define, Deliver and Drive. Discover diagnoses the gaps. Desire aligns the leadership vision with the growth ambition. Define designs the systems, rhythms and pathways. Deliver embeds development in the business through coaching, experiences and rituals. Drive builds leaders who grow other leaders. Coaching belongs mostly in Deliver. That is why a diagnosis comes first: it tells you whether the thing to deliver is coaching at all. You can read the full model on my 5D Leadership Framework page, and my leadership assessments measure a group of leaders on the same instruments.

For leaders working through a diagnosis, I built CapabilityAI to help apply the 5D model to a specific challenge between conversations. I mention it as my own product. It supports reflection and does not replace a diagnosis led by a person who can see your organisation.

What a 30-day diagnosis looks like

1. Week one: define the problem in business terms: Write one paragraph that a director could read: what is happening, where, since when, and what it costs. Remove adjectives about people. Name the behaviour and its effect.
2. Week two: gather evidence from several angles: Interview the leader's direct reports, peers and sponsor separately. Trace three recent decisions from start to finish. Review the incentives and the role's mandate. Where a 360 exists, read it; where it does not, consider one.
3. Week three: sort what you found: Place each finding in a pillar: alignment, identity, cadence, pathway or culture. Mark each as a person issue, a system issue or both. Look for patterns across leaders.
4. Week four: decide and brief: Choose the intervention, name the sponsor, write the brief, agree the measures and set a review date. If coaching is the answer, the brief now states what the coach should work on and what lies outside their remit.

Keep the diagnosis proportionate. For a single leader in a stable team, a few conversations and a decision trace may be enough. For a pattern that appears across several leaders, spend the full month and involve someone outside the line of command, because people describe their own organisation more candidly to an outsider. Whatever the scale, write the findings down in a page, share them with the leader concerned, and record what you decided and why. That record becomes the baseline against which you later judge whether the intervention worked.

How a CHRO should brief the CEO before buying coaching

A chief people officer is often the person who sees the pattern first and the one asked to arrange the coaching. The most useful thing they can do is slow the request by a few weeks. Say to the CEO: I will arrange the right support, and first I want to establish what is driving this, because coaching a leader inside a faulty system risks making things worse. Offer the 30-day diagnosis. Ask the CEO to name what they would expect to see change in the business, not only in the leader.

Then make the CEO's own part visible. In many cases the executive team's difficulty includes the CEO's behaviour and the decision rights the CEO has set. A CHRO who can raise that early and calmly does the business a service. My article on what makes executive coaching for CEOs effective covers how to approach it, and executive coaching or leadership systems frames the choice for the board.

What boards should ask before approving a coaching budget

• What problem is the coaching meant to solve, and how do you know the problem is the person and not the structure?
• What evidence did you gather, from whom, before choosing this intervention?
• Who sponsors the work and what will they see?
• What will be different in 90 days, and how will we know?
• If the diagnosis showed a system issue, what would you do instead?

If you want a view on the wider field, my strategic handbook for executive leadership coaching and the overview everything you need to know about executive leadership coaching are good places to begin.

When executive coaching is exactly right

None of this is an argument against coaching. When the diagnosis points to the person, coaching is often the best tool available. That happens when a leader has moved into a role that needs a different identity and wants help making the move, when several independent sources confirm a blind spot the leader has not seen, when a leader is performing well in a stable system and is ready for a harder challenge, or when the leader needs a confidential place to think through a decision. In those cases the system is sound and the person is the lever. A good diagnosis lets you say so with confidence, and it gives the coach a brief that can be followed and reviewed.

The distinction I would keep

Coaching develops the leader; diagnosis decides whether the leader is the thing to develop. The order matters, because a leader who has been coached for a problem the system created has been given an answer to the wrong question.

If you are weighing a coaching decision now, start with the Leadership Diagnostic or see how I work with CEOs and CHROs.