Two engineers report to the same manager. Same week, same task: own the migration of a creaking billing service. One has shipped three migrations and is quietly bored; the other has never owned anything this size and is hiding it well. Hand them both the same memo, "you've got this, ping me if you're stuck", and one will fly while one quietly drowns. The leadership didn't fail because the style was bad. It failed because it was identical.

The quick version

  • There is no single best leadership style, anyone selling you one is selling.
  • The competent move is to read the situation and the person, then flex your style to fit.
  • The real failure mode is using one style for everything, to the leader with one style, every problem looks the same.
  • So the practice is simple to state, hard to do: diagnose this person's readiness for this task, pick the matching style, and re-diagnose as it moves.

The idea in depth

The models you have heard of, situational, servant, transformational, Goleman's six, are not rival religions where you pick one and convert. They are lenses: vocabularies for noticing what a moment needs. Used well, together, they sharpen one question the single-style manager never asks, what does this person, on this task, right now, actually need from me? Here are the four that earn their keep, each with the move it tells you to make on Monday.

Goleman's six styles: a leader's weather report

The most useful inventory of leadership styles came out of data rather than a guru's say-so. In "Leadership That Gets Results" (Harvard Business Review, March 2000), Daniel Goleman drew on a study of more than 3,000 executives, conducted with the consultancy Hay/McBer, to map which leadership behaviours actually moved organisational climate and performance. He found six distinct styles, each rooted in a facet of emotional intelligence:

  • Coercive (commanding), "do what I tell you." Demands immediate compliance.
  • Authoritative (visionary), "come with me." Mobilises people toward a vision.
  • Affiliative, "people come first." Builds harmony and emotional bonds.
  • Democratic, "what do you think?" Builds buy-in through participation.
  • Pacesetting, "do as I do, now." Sets high standards and expects self-direction.
  • Coaching, "try this." Develops people for the future.

Goleman's headline finding is the one most people skip: the styles with the most consistently positive effect on climate were the authoritative, affiliative, democratic and coaching ones, while coercive and pacesetting, used as a default, tended to poison the weather. The deeper finding is that the best leaders don't pick a favourite. They use most of the six in a given week, fluidly, "like a pro golfer pulling a club from the bag", Goleman's image, choosing by the shot in front of them.

So the move is: stop asking "what kind of leader am I?" and start asking "what does this moment need?" A production outage at 2am needs the coercive register, clear orders, no debate. The same register in a Monday retro will gut the team. Name your own default style (most of us have one or two we reach for under pressure) and treat the others as clubs you have under-practised, not personality traits you lack.

flowchart LR
  A(["What does THIS moment need?"]) --> B(["Crisis or genuine emergency?"])
  B -- yes --> C(["Coercive: clear orders, fast"])
  B -- no --> D(["Team lost the bigger why?"])
  D -- yes --> E(["Authoritative: set the vision"])
  D -- no --> F(["Trust frayed, people hurting?"])
  F -- yes --> G(["Affiliative: repair the bonds"])
  F -- no --> H(["Need buy-in or good ideas?"])
  H -- yes --> I(["Democratic: ask, then decide"])
  H -- no --> J(["Coaching: grow them for next time"])
A rough decision aid for selecting a style, the everyday register, with coercive reserved for true emergencies and pacesetting used sparingly. After Goleman, "Leadership That Gets Results" (HBR, 2000). Leaders Loop

Situational leadership: match the style to readiness

Goleman tells you the styles exist and that good leaders mix them. The obvious next question, mix them how?, is what Paul Hersey and Ken Blanchard set out to answer with Situational Leadership, the most widely taught model in corporate training rooms. Its premise is intuitive: the right style depends not on you but on the readiness of the person for the specific task.

The model crosses two behaviours, how much direction you give (task instruction) and how much support you give (encouragement, involvement), into four styles, and matches each to a development level of the follower:

  • D1, enthusiastic beginner (low skill, high will) → S1 Directing: high direction, low support. Tell them exactly what and how.
  • D2, disillusioned learner (some skill, low will) → S2 Coaching: high direction and high support. Keep steering, but rebuild confidence.
  • D3, capable but cautious (high skill, variable will) → S3 Supporting: low direction, high support. Hand over the wheel, stay in the passenger seat.
  • D4, self-reliant achiever (high skill, high will) → S4 Delegating: low direction, low support. Set the outcome and get out of the way.
quadrantChart
  title Situational Leadership: style by follower readiness
  x-axis "Low directive" --> "High directive"
  y-axis "Low supportive" --> "High supportive"
  quadrant-1 "S2 Coaching (D2)"
  quadrant-2 "S3 Supporting (D3)"
  quadrant-3 "S4 Delegating (D4)"
  quadrant-4 "S1 Directing (D1)"
  "Enthusiastic beginner D1": [0.78, 0.22]
  "Disillusioned learner D2": [0.80, 0.80]
  "Capable but cautious D3": [0.22, 0.80]
  "Self-reliant achiever D4": [0.20, 0.20]
The situational grid: directive behaviour (horizontal) against supportive behaviour (vertical), with the four styles S1–S4 mapped to development levels D1–D4. After Hersey & Blanchard. Leaders Loop

The trap the model warns against is the most common management error there is: treating readiness as a property of the person rather than the task. Your strongest engineer is a D4 on the codebase she wrote and a D1 the first time you ask her to manage a vendor negotiation. Readiness is task-specific and it moves. So does the style you owe her.

One honest caveat, because the model's popularity outruns its proof. Situational Leadership feels right and is teachable in an afternoon, none of which makes it true. In his much-cited critical review, Claude Graeff (1997, The Leadership Quarterly) found the theory beset by internal logical inconsistencies and weak conceptual grounding, and decades of studies since have produced, at best, partial support. The broader picture is summarised in the academic literature: a gap between popularity with practitioners and thin validation in research. The same skepticism applies across the field, these are lenses, not laws. Treat them as vocabularies for noticing, which is genuinely valuable, and resist anyone who sells them as physics.

So the move is: before your next delegation, score the person on this one task, not in general, on two axes: can they (competence) and will they without hand-holding (commitment). Then deliberately set your direction and support to match. Most over-management is S1 aimed at a D4; most "they dropped the ball" is S4 aimed at a D1.

The two engines: transformational, transactional, and the servant stance

Goleman and Hersey–Blanchard are about how you behave in the moment. A partly orthogonal question is what you move people with, the domain of the Full Range Leadership model developed by Bernard Bass and Bruce Avolio (building on James MacGregor Burns's 1978 distinction), measured by their Multifactor Leadership Questionnaire.

Transactional leadership runs on exchange: clear expectations, contingent reward, correction when things slip, the management of the deal. Transformational leadership runs on meaning: a compelling vision, inspiration, intellectual stimulation, and individualised attention to each person's growth. Of the styles in this literature, transformational leadership has the strongest and most replicated association with effort, satisfaction and effectiveness, and, unlike the situational model, that is a finding with a substantial peer-reviewed base behind it. The honest reading is that you need both: transactional foundations are the floor (without them, "vision" is just a poster); transformational behaviours are the ceiling (what turns a competent team into one that does its best work when you are not in the room).

The last lens is less a style than a stance. In his 1970 essay The Servant as Leader, Robert K. Greenleaf, a retired AT&T executive, not an academic, coined servant leadership and put the question that reframes all the others: who is the leadership for? As the Greenleaf Center still treats as foundational:

"The servant-leader is servant first… It begins with the natural feeling that one wants to serve, to serve first. Then conscious choice brings one to aspire to lead." , Robert K. Greenleaf, The Servant as Leader, 1970

Read servant leadership not as a seventh style to slot beside the others, but as the posture from which you choose them. The directive S1 you give a frightened beginner, the vision you set, the coercive order in the outage, all can be delivered to serve the person and the work, or to serve your own standing. (As with the situational model, the empirical evidence here is still maturing; its strongest claim is ethical, not predictive.)

So the move is: audit your last month. All status and corrections (transactional), no meaning? Have one conversation this week about where a person is heading, not where the project is. And before any hard conversation, ask one question, "am I about to do this for them and the work, or to them?" The honest answer usually rewrites your opening sentence.

A worked example

Back to the billing migration and the two engineers. A manager who has internalised the diagnostic, rather than a favourite style, runs two playbooks for the same task.

With Priya, who has shipped migrations before, the task reads D4. The instinct might be to over-prepare a detailed brief; the diagnosis says don't. The move is S4 delegating in a transformational frame: "This is yours. The outcome is zero billing errors through the cutover and a runbook the next person can follow. I trust your judgement on the how, and owning this end-to-end is exactly the kind of work that gets you to staff engineer." One sentence of outcome, one of trust, one of meaning. Then silence, and availability.

With Marco, who is hiding that he has never owned anything this size, the same task reads D1, even though he is a strong engineer elsewhere. Hand him the Priya speech and you have set him up to fail and call it his fault. The move is S2 coaching, high direction and high support: "Let's break this into the first three steps together now, and I want a fifteen-minute check-in at the end of each day this week, not because I doubt you, but because this is a big one and I'd do the same in your shoes." High structure, framed as normal, not remedial. As Marco's competence climbs over the next fortnight, the manager deliberately dials direction down toward S3, then S4, flexing within the same project as the readiness moves.

Same manager, same task, same week, two styles, chosen on purpose. That is the entire skill. Not having the right style. Having more than one, and the diagnostic to know which.

Frequently asked questions

Isn't deliberately switching styles inauthentic?

It is the right thing to worry about, and the answer is no. Flexing style is not faking emotion; it is choosing behaviour. A parent is gentle with a scared child and firm at a kerb without being "inauthentic" in either, both are real, the situation selects which fits. The fake version is warmth deployed as manipulation, or vision as theatre. Authenticity lives in the intent (Greenleaf's servant-first posture), not in refusing to adapt. The leader who insists "this is just how I am" is usually defending a comfort zone and charging the team for it.

Which leadership style is best?

None of them, that is the whole point. There is no single best style; there is only the style that fits this person, this task and this moment. The leaders Goleman studied who lifted climate most were the ones who commanded the most styles and switched between them fluidly, not the ones with the "right" favourite.

What's the difference between transformational and servant leadership?

Transformational leadership (Bass & Avolio) is a set of behaviours, vision, inspiration, intellectual stimulation, individual attention, aimed at lifting performance and effort, and it has a strong empirical base. Servant leadership (Greenleaf) is a stance, a prior question about whom the leadership serves. You can behave transformationally from either a self-serving or a servant-first place; servant leadership is the posture, transformational the technique.

Does situational leadership actually work?

As a diagnostic habit, yes, it makes you ask "how ready is this person for this task?" before you delegate, which is valuable on its own. As a validated scientific theory, the record is thin: Graeff's 1997 review found logical inconsistencies and weak grounding, and later studies offer only partial support. Use it as a lens that sharpens the question, not a law that guarantees the answer.

How do I choose, in the corridor, before a delegation?

Run a four-beat check. Diagnose the moment (crisis buys the coercive register; almost nothing else does). Diagnose the person for this task (can they? will they without hand-holding? that sets your S1–S4 dial). Check the engine (is the transactional floor in place before you add transformational meaning?). Check the stance (am I doing this for them, or to them?). Then flex as competence and confidence move, within the project, not just between jobs.

Where to go next

For the engine that turns a flexed style into discretionary effort, see the sibling tool on motivating and inspiring teams; and for a portrait of one leader who built a career on this kind of deliberate, context-first judgement, read our profile of Rob Alford.

Related in the Toolkit