Coaching vs Directing: What Managers Need to Know

In 2000, Daniel Goleman published “Leadership That Gets Results” in Harvard Business Review, describing six distinct leadership styles and their measurable effect on organizational climate. One finding has aged particularly well: the coaching style had a reliably positive effect on climate and performance, and it was the style leaders used least. Goleman’s explanation was not that leaders lacked the skill. It was that coaching takes longer in the moment, and managers under pressure reach for the faster option.

That tradeoff is the whole subject. Directing is faster now. Coaching is faster later. Most managers know this and still default to directing, because “now” is when the pressure is. The useful question is not which approach is better in the abstract. It is how to tell, in a specific situation, which one the moment actually calls for.

What the Research Shows

Coaching Is Underused, Not Unknown

Goleman’s data across thousands of executives found coaching to be the least frequently used of the six styles despite its positive climate effect. Nearly two decades later, Herminia Ibarra and Anne Scoular made a related observation in “The Leader as Coach”: when managers are asked to coach, many produce something closer to advice-giving with a question mark attached. They ask questions to which they already have the answer and steer toward it. The intent is coaching. The mechanics are directing.

The Right Choice Depends on the Situation, Not the Person

The situational leadership framework developed by Paul Hersey and Ken Blanchard makes the point most directly: the appropriate level of direction varies with the person’s competence and confidence for the specific task at hand, not with their overall seniority. A capable senior engineer facing an unfamiliar regulatory process needs direction. A junior analyst working in the area they know best needs room. Treating direction as a property of the person rather than the task is one of the most common management errors, and it reliably produces both under-supported new hires and over-managed experts.

Directing is faster now. Coaching is faster later. Most managers know this and still default to directing, because “now” is when the pressure is.

Coaching Requires the Conditions to Exist First

Coaching assumes the person can think out loud, including about their own uncertainty. That is a psychological safety precondition. In a team where admitting you do not know something carries a cost, coaching questions are experienced as tests rather than support, and people respond by guessing at the answer the manager wants. We covered the underlying dynamic in psychological safety as infrastructure. Without it, coaching does not merely fail. It reads as evasion, as though the manager knows the answer and is withholding it.

Why This Matters for Teams

The costs of getting this wrong run in both directions, and they look nothing alike.

Over-directing produces capable people who cannot operate without checking. Every decision routes upward, because that is what the pattern has taught. The manager becomes the constraint on the team’s throughput and usually experiences this as being indispensable rather than as a design flaw. The tell is simple: when the manager takes a week off, does work slow down? If it stops entirely, the team has been directed into dependency.

Over-coaching is less discussed and equally real. A manager who responds to every question with a question, including questions with a straightforward factual answer, is not developing anyone. They are withholding. When someone asks which of two systems is the system of record, they need to be told. Turning that into a developmental moment wastes their time and signals that asking you anything will cost twenty minutes, which is an efficient way to stop being asked.

There is also a speed cost that gets ignored in the enthusiasm for coaching. In a genuine emergency, direction is correct and coaching is negligent. During an outage, nobody should be asked what they think the options might be. The skill is not preferring one mode. It is switching cleanly and signalling which mode you are in, so the other person knows whether they are being consulted or instructed.

The Gap the Data Reveals

The research establishes that coaching works and that managers under-use it. What it does not supply is the decision rule, and the absence of one is why “manage more like a coach” tends to produce either mimicry or nothing.

The gap is that most training teaches coaching as an identity rather than a mode. Managers are told to be a coach, which is unfalsifiable and unhelpful on a Tuesday afternoon with four decisions pending. What they need is a rule for choosing, applied per situation, several times a day. Identity-based framing also makes switching feel like a failure of consistency, so managers who have committed to being coaches stay in coaching mode during moments that plainly call for a decision.

The second gap is that coaching capacity is treated as free. It is not. Coaching a decision takes perhaps three times as long as making it, and that time comes out of the same week as everything else. A manager with nine direct reports and a full delivery load does not have the hours to coach every decision, and telling them to do it anyway produces guilt rather than behavior change. The honest version is that coaching is an investment allocated deliberately, which means choosing which situations get it.

The Gartner data on leadership capability investment shows organizations pouring money into manager development while the structural conditions, span of control and workload, go untouched. That is the shape of the gap. Skill training against a capacity constraint.

What This Looks Like in Practice

A workable rule fits in a sentence: direct when the cost of being wrong is high and the person lacks context, coach when the cost of being wrong is recoverable and the person will face this class of problem again.

Sort by Reversibility, Not Importance

The instinct is to direct on important things and coach on trivial ones, which gets it exactly backwards a lot of the time. Importance is not the variable. Reversibility is. A high-stakes but reversible decision is an excellent coaching opportunity, because the learning is proportional to the stakes and the downside is recoverable. A low-stakes irreversible decision, such as anything touching a customer’s data or a public commitment, deserves direction regardless of how minor it seems.

Say Which Mode You Are In

Announce it. “I’m going to just tell you this one, we don’t have time to work through it.” Or: “I have a view here but I want to hear how you’d approach it first.” Two seconds of framing eliminates the most common failure in both directions, which is the employee guessing whether they are being asked or told. Ambiguity about mode is worse than either mode.

Give the Answer, Then Reconstruct It

When time forces direction on something the person should eventually own, direct now and come back to it. Later, in the one-on-one, walk through how you got to that answer and what you were weighing. This captures most of the developmental value without paying the in-the-moment cost. It is the highest-leverage habit on this list and almost nobody does it, because once the problem is solved the urgency evaporates.

Watch the Escalation Pattern

Track what gets escalated to you over a month. If the same category keeps arriving, that category is a coaching target and directing it again will not change anything. If escalations are genuinely varied and novel, your team is well calibrated and you should keep directing the one-offs. The pattern in the escalation log is a better diagnostic than any assessment of individual readiness.

Goleman’s original finding still holds, and the reason has not changed. Coaching remains the underused style because its return arrives after the quarter in which you paid for it. Directing clears today’s queue. Coaching changes what is in the queue three months from now. Both are correct choices at different moments, and the managers who are actually good at this are not the ones who coach the most. They are the ones who can tell which situation they are in before they open their mouth.

Related Reading

• How to Run One-on-Ones That Actually Change Outcomes
• Radical Candor at Scale: Building Feedback Systems That Work Across Teams
• Psychological Safety Is Infrastructure, Not Culture
• Gallup’s Manager Research: Why Managers Account for 70% of Engagement Variance
• The Feedback Script That Prevents Surprise Performance Reviews