I've sat across the table with a lot of executives who describe their managers the same way: good operators, bad coaches. They know the function. They can hit the number. Ask them to grow someone else's capability, though, and most freeze, hand it to HR, or simply stop trying. That's not a training gap you patch with a single workshop. It's the coaching capability shortage, and it is bigger, and more expensive, than most boards realise.
The pattern is consistent across the companies I work with. Each invests heavily in what its managers produce and almost nothing in what those same managers develop in other people. Performance reviews measure output. Promotions reward individual delivery. Nobody sits down and asks whether a manager can actually grow the capability of the six or ten people reporting to them. The organisation assumes coaching happens by osmosis: that competent people naturally know how to build competence in others. They don't. Coaching is a distinct skill. Most managers were never taught it, never assessed on it, and never held accountable for it.
The gap is easiest to see at the promotion moment. Someone gets moved into management because they were excellent at the job underneath it: the best salesperson becomes the sales manager, the sharpest engineer becomes the tech lead. Nothing about that promotion tests, teaches or even asks whether they can develop capability in someone else. The organisation hands them a team and assumes the coaching will follow the title. It rarely does, because the skill that got them promoted and the skill the new role actually requires are almost entirely different.
What the coaching capability shortage actually looks like
Gartner's own research puts a number on what most leaders already sense. In a 2019 survey covering the skills gap in corporate learning, Gartner found that 45% of managers don't feel confident in their ability to develop the skills employees need today. Not the skills employees needed five years ago. The skills they need today, in their current role, this quarter. Nearly half the management layer at a typical organisation is unsure it can do the one job that keeps the rest of the workforce capable.
Confidence is only half the problem. Gartner also found that managers spend on average just 9% of their time on developing their direct reports. Run the maths on a standard working week and that is roughly half a day a fortnight, spread thin across every person on the team, competing with every deadline, every fire and every meeting that actually gets defended on a calendar. Coaching loses that fight almost every time, not because managers do not care, but because nothing in the system protects the time or rewards the effort.
The tension shows up clearest at the top. More than 70% of HR executives believe managers should be more involved in coaching employees than they were three years ago, according to the same Gartner data. Senior leadership already knows the answer. It just has not built anything that makes the answer possible. Wanting managers to coach more and equipping them to coach well are two entirely different programmes, and most organisations only run the first one.
45%: of managers lack confidence to develop the skills their people need: Gartner, 2019
9%: of a manager's time spent developing direct reports, on average: Gartner, 2019
70%+: of HR executives want managers more involved in coaching: Gartner, 2019
Why coaching gets treated as a soft extra
Most organisations sort skills into two buckets: the hard ones that show up on a P&L, and the soft ones that show up on an engagement survey. Coaching gets filed in the second bucket, which means it gets funded like the second bucket. A day of training here, a values poster there, a line in the leadership competency framework that nobody reads past the first review cycle. Meanwhile the capability that actually determines whether a strategy survives contact with the org chart sits unmanaged, undermeasured, and largely unbuilt.
The mistake is definitional, not moral. Coaching is not a personality trait some managers happen to have. It is a technical skill built from specific behaviours: asking a real question instead of handing over the answer, holding a silence long enough for someone to think, giving feedback that is precise enough to act on, knowing when to step back and let a person fail small so they do not fail large later. Every one of those behaviours can be taught, practised and measured. Almost none of them are, at scale, in most companies.
That is the reframe worth sitting with. It's not a soft skill: it's a structural capability, and structural capabilities are engineered, not absorbed. You do not get a reliable supply of managers who can grow other people's competence by hoping the good ones figure it out. You get it by building the capability the same way you would build any other operating system inside the business, deliberately, with an owner, a standard and a feedback loop.
The structural advantage of building it deliberately
The shift is visible the moment an organisation treats coaching as infrastructure rather than an HR nicety. Every manager becomes a multiplier instead of a ceiling. A team led by someone who can genuinely develop capability in others compounds: skills transfer faster, succession gaps close earlier, and the business stops depending on a handful of irreplaceable individuals to hold the whole operation together. The best managers I have studied treat developing talent as a core deliverable of the role, not an occasional favour they do for a promising junior.
Competitors that never build this capability are not just weaker at retention, although they usually are that too. They are running with a permanent tax on every strategic initiative that depends on people executing differently than they did last year. New systems, new markets, new operating models all require the workforce to acquire new capability quickly. If the manager closest to that workforce cannot coach the change into people, the change stalls at the front line no matter how well it was designed at the top.
There's a retention cost too, though it usually gets tracked as an unrelated line item. People do not leave capable teams as often as they leave managers who cannot help them grow. Exit interviews call it 'lack of development opportunities' and HR treats it as a perks problem, more training budget, more internal mobility postings, when the actual cause sits one level up: nobody taught the manager how to develop the person in the first place.
How the coaching layer fits inside a leadership capability architecture
I use a simple structure to make coaching capability buildable rather than aspirational: a designed layer with its own inputs, standards and outputs, not a separate wellbeing initiative bolted onto the org chart.
- Diagnose: Assess coaching behaviour directly, not proxy metrics like engagement scores. Watch how a manager actually runs a development conversation before assuming they can.
- Design: Build coaching into the management role itself: protected time, a defined cadence, and a standard for what a good development conversation contains.
- Develop: Teach the specific behaviours, questioning, feedback, delegation with stretch, rather than a generic soft-skills course that never touches the manager's real team.
- Sustain: Measure whether capability in the team is actually growing, and hold managers accountable for that outcome the same way you hold them accountable for output.
What it takes to build the coaching layer into management
Building this capability is a design problem before it is a training problem. Training without redesigning the role just produces managers who know the theory and have no time, no incentive and no standard forcing them to use it. The organisations that actually close the gap change four things at once.
- Protect the time: Put a fixed weekly slot for development conversations on every manager's calendar and treat it with the same seriousness as a customer meeting. Time that is optional gets cancelled first.
- Define the standard: Write down what a good development conversation actually contains: a specific behaviour observed, a question that surfaces the person's own thinking, and one concrete commitment before the next check-in.
- Assess the behaviour: Test coaching skill directly, through observed conversations or structured assessment, rather than inferring it from tenure or technical performance. A brilliant individual contributor is not automatically a capable coach.
- Build accountability into the role: Make the capability growth of a manager's team a measured part of that manager's own performance, not an aspiration mentioned once a year in a review.
- Coach the coaches: Give managers their own coaching, from someone senior enough to model the behaviour rather than just describe it. Capability is easier to copy than to invent from a slide deck.
Each of these four moves is simple on its own. What is rare is doing all of them together, deliberately, rather than picking whichever is cheapest to run this quarter. Coaching leadership style is often treated as a personality preference some managers happen to have. Treated as infrastructure instead, it becomes something you can specify, teach and measure like any other operating capability. The same logic applies to the practical craft of the conversation itself: the specific techniques that make a coaching conversation actually land are learnable in weeks, not years, once someone is willing to teach them properly.
Where the coaching capability shortage gets worse, not better
The shortage is not staying flat. Every wave of technology change asks managers to develop new capability in their teams faster than the last one did, and the current wave is the fastest yet. Leadership pipelines are already showing strain under AI-era pressure, and coaching sits directly in the middle of it. If managers could not reliably develop people before the pace of change accelerated, they are not going to manage it better once the skills employees need are shifting every two quarters instead of every two years.
This is where the competitive edge argument gets sharp rather than theoretical. Two companies can buy the same tools, hire from the same talent pool and copy the same strategy. What they cannot copy quickly is a management layer that already knows how to grow capability in people under pressure. That takes years to build if you start from nothing, which means the organisations starting now are the ones that will have it when the gap between winners and laggards actually opens up.
The distinction worth taking into your next leadership meeting
A confident, well-resourced manager who cannot coach will still hit this quarter's number. A manager who can coach builds a team that keeps hitting numbers after that manager has moved on. One produces results. The other produces an organisation. Most leadership teams are optimising for the first outcome and hoping the second one shows up on its own. It will not. Coaching capability is not a soft asset that appears when the culture is healthy enough. It is a specific, buildable, measurable skill, and the gap between the companies that build it and the companies that assume it is one of the clearest predictors of which one is still standing when the next disruption hits.
If you want to see what building this capability properly looks like inside a real leadership team, Capability AI gives managers structured, ongoing support for exactly this kind of development work, and the Architecture Accelerator programme is where I take leadership teams through building coaching capability into the management layer itself, rather than bolting it on as a course.