We spend most of our working lives arguing that problems people blame on individuals are usually caused by systems. Deming’s point — that the system accounts for most of the variation in performance — has held up remarkably well.
But it isn’t universal, and pretending otherwise does businesses a disservice. Leadership coaching for businesses exists precisely for the cases the system explanation doesn’t cover.
Sometimes the process is fine, the tools are fine, the people are capable, and the results still aren’t there. At that point you’re looking at a leadership problem.
How to tell the difference
The distinction matters commercially, because the two problems cost very different amounts to fix and running the wrong intervention wastes both money and credibility.
Here’s the test we’d apply.
If the same process produces different results under different managers, it’s a leadership problem. Two teams, same system, same targets, materially different outcomes. The system is capable. Something about how it’s being led isn’t.
If the same process produces poor results under every manager, it’s a process problem. You can rotate good people through it indefinitely and it will defeat all of them.
That single question resolves most cases. It’s worth asking before commissioning either kind of work.
The symptoms that point to leadership
A few patterns come up repeatedly.
Improvement work that doesn’t stick. A process gets redesigned, works for a quarter, then drifts back. In our experience this is rarely a flaw in the redesign. It’s an absence of anyone holding the new way in place — which is a management behaviour, not a process feature.
Information that doesn’t travel upwards. Problems are known at the front line and unknown at the top. Usually this means somewhere in the middle, raising problems has been unrewarding.
Capable people leaving. People rarely leave jobs they’re good at because of processes. Turnover concentrated under one manager is data.
Decisions that don’t get made. Work stalls waiting for a call that nobody will make. That’s often less about capability than about a culture where being wrong is more costly than being slow.
Everything escalating. If a manager’s team can’t resolve anything without them, they’ve either not built the capability or not delegated the authority.
Why this happens to good people
Most managers in mid-sized UK businesses were promoted because they were good at the job below.
The best quantity surveyor becomes the commercial manager. The most reliable site manager becomes the contracts manager. The strongest technician runs the team.
None of those promotions come with training in the thing that’s now the actual job — getting results through other people, having difficult conversations, holding standards without micromanaging, developing someone else’s capability rather than doing the work yourself.
We promote people for being excellent at a job, then give them a completely different one and wonder why they struggle.
This isn’t a criticism of those managers. It’s a criticism of how businesses handle the transition. The skill gap is entirely predictable and almost always unaddressed.
What the evidence suggests about priorities
There’s a finding in the RICS Construction Productivity Report 2026 that’s relevant well beyond construction.
Asked which interventions would most improve productivity, respondents rated upskilling people highest — an average of 47% rating it high-impact, and the top-ranked intervention in four of the five regions surveyed. Automation, by contrast, was rated high-impact by only 17% of UK respondents, the lowest confidence of any region.
Nearly 3,000 professionals, and their collective view is that the constraint is capability rather than technology.
We’d read that as a reasonable reflection of experience. Businesses that have invested in systems and seen limited return often find the missing ingredient was whether anyone was equipped to lead the change.
What coaching actually does
Leadership coaching gets confused with a few things it isn’t.
It isn’t training. Training transfers knowledge to a group. Coaching works on one person’s specific situation, with their actual team and their actual problems.
It isn’t therapy, and it isn’t a performance management route. Using coaching as a step towards exit poisons it for everybody else in the business.
It isn’t advice, mostly. The value is less in being told what to do and more in having a structured, external space to think — with someone who has no stake in the internal politics and will ask the question colleagues won’t.
In practice it usually works on a small number of concrete things: how this manager handles conflict, how they delegate, how they hold people to a standard without doing the work themselves, and how they communicate change to a team that has seen initiatives come and go.
When to choose coaching over training
Both have their place, and the choice is fairly simple.
Choose training when several people need the same capability. Delegation, difficult conversations, running effective one-to-ones — these are teachable in a group and cheaper that way.
Choose coaching when one person is in a specific situation that generic content won’t address. A newly promoted manager struggling with former peers. A capable leader whose team keeps missing targets for reasons nobody can name. A senior person who has plateaued.
Many businesses need both — training to lift the baseline across a management team, coaching for the two or three individuals whose performance disproportionately affects the business.
What a coaching engagement actually looks like
Because this varies so much between providers, it is worth being specific about the shape of the work.
Duration. Meaningful change in how someone leads takes months, not weeks. Six sessions over six months does more than six sessions over six weeks, because the value sits in what happens between them.
Cadence. Fortnightly or monthly. The gap is where the manager tries something, it goes imperfectly, and that becomes the material for the next conversation.
Focus. Two or three specific things, agreed at the start. Coaching that tries to address someone’s entire leadership style produces pleasant conversations and no change.
Confidentiality. The content stays between coach and coachee. What can reasonably be shared with the business is whether the engagement is on track and whether the agreed goals are moving — not what was said.
Measurement. This gets neglected. Agree at the outset what would be observably different in six months: fewer escalations, a specific team metric, retention within the team, a behaviour their own manager would notice. Vague goals produce unprovable outcomes.
Getting the business to support it
One practical point. Coaching a manager while leaving the conditions around them unchanged rarely works.
If someone is being coached on delegation while their own manager continues to demand involvement in every decision, the coaching loses. The person above needs to know what is being worked on and to change their own behaviour accordingly — not the content of sessions, but the direction.
This is why we usually ask for a short conversation with the coachee’s manager at the start and at the midpoint. Not to report, but to make sure the environment supports what is being built.
The honest caveat
Coaching is not a fix for a badly designed job, an impossible target, or a process that defeats everybody who touches it.
If a manager is failing because the system they’re operating is broken, coaching them is unkind as well as ineffective. It communicates that the problem is them when it isn’t.
That’s why the diagnostic question at the top matters. Get it wrong in that direction and you damage a capable person while leaving the actual problem untouched.
We do both kinds of work, and a fair amount of our time goes on establishing which one a business actually needs before anybody commits to either.
Find out about our leadership coaching → /coaching-mentoring work, or read how we approach business process improvement → /business-improvement when the system is the constraint.