Domain 1 — People Leadership
People Leadership is the work of building, growing, and retaining effective engineers, teams — and, at altitude, leaders. It is the most human part of the job and, at every level, the part where the previous level's strengths become liabilities: the instincts that made you a strong engineer (solving problems yourself) are nearly the opposite of what makes a strong team leader, and the instincts that made you a strong team leader (being close to every person and every problem) are nearly the opposite of what makes a strong leader of managers. This domain carries 20% of every exam (12 questions on EMA-I, 15 on EMP-II, 18 on EME-III).
The altitude arc of this domain is the clearest in the framework: at Associate you lead people; at Professional you lead the people who lead people; at Expert you build the system that produces and sustains leaders. The throughline is judgment about people under uncertainty — who to trust with what, what feedback to give and when, whom to hire and promote, how to handle underperformance, and how to invest in growth. None of these have formulaic answers; the exams reward the response that balances care for the individual with responsibility to the team, the organisation, and the work.
1.1 Delegation and ownership
Delegation is how a manager scales beyond their own two hands — and how the people around them grow. The core skill at every altitude is matching the scope of what you hand off to the person's current capability plus a deliberate stretch: enough challenge to develop them, not so much that they fail unsupported.
A useful mental model is task-relevant maturity: how much direction someone needs depends not on their seniority in general but on their experience with this specific kind of work. A senior engineer may need close support on their first incident command even though they need none on system design — and a brilliant engineer newly promoted to manager has low task-relevant maturity at management itself, however senior they are technically. This connects to situational leadership — flexing between directing, coaching, supporting, and delegating as competence and confidence grow.
Delegation is not abdication. You delegate the work and the decisions, but you retain accountability for the outcome — this remains true whether the thing delegated is a task, a team, or a division. Good delegation is explicit about three things: the desired outcome, the constraints (deadline, budget, what not to touch), and the level of authority ("decide and proceed" vs. "recommend and check with me"). What changes with altitude is the object of delegation: tasks and projects at Associate, whole teams at Professional, and organisational structure itself at Expert.
At Associate — delegating work
The first-line version is delegating meaningful work to engineers: outcomes rather than steps, a stretch with a safety net, explicit decision rights, and the discipline not to take the work back the moment it gets messy. Rescuing at the first sign of struggle teaches the team they can't be trusted; hoarding the interesting decisions and delegating only the boring work teaches them delegation is a chore, not an investment.
Strong judgment looks like: delegating outcomes rather than dictating steps; giving someone a stretch with a safety net; being clear about decision rights; resisting the urge to take a task back when it wobbles.
Common pitfalls: delegating only the boring work; "delegating" by dumping an under-specified task and disappearing; swooping in to rescue prematurely; delegating authority you then quietly override.
At Professional — delegating teams
A manager of managers delegates teams. The discipline is identical — outcomes, constraints, decision rights, retained accountability — but the temptations are stronger, because you used to do the manager's job and could often do it faster. The characteristic failure at this altitude is reaching around your managers: overriding their decisions with their engineers, fixing their team's problems directly, or letting their escalations quietly become your workload. A manager who escalates nearly every cross-team decision to you, or whose 1:1s with their own team have quietly stopped, is struggling at scale — the response is coaching and explicit expectations, not silently absorbing their job.
Two recurring Professional situations test this competency directly. First, the new manager who micromanages — reviewing every pull request, redoing junior work — while their team's growth stalls. The effective intervention is not to strip their duties or send them back to an IC role at the first stumble: coach them to manage outcomes rather than tasks, help them define what to delegate and what "good enough" looks like, and protect them while they adjust. Second, span of control: how many managers or teams one person can effectively lead is determined not by a magic number but by the maturity of the managers, the stability of the work, and how much cross-team coordination each report requires.
Strong judgment looks like: handing each manager a genuine mandate with explicit decision rights between you and them; using skip-levels for signal, never for decisions; helping an overloaded manager still doing significant IC work to shed it before they burn out; coaching a struggling manager's judgment rather than doing their job.
Common pitfalls: taking a team back when it struggles instead of developing its manager; managing your managers' engineers directly; treating every escalation as yours to solve; tolerating a manager whose personal heroics mask a team that can't operate without them.
At Expert — delegating through structure
At organisational scale, ownership is designed, not assigned conversation by conversation. The org structure, team charters, and decision rights determine who owns what — and a reorganisation is, above all, a redesign of ownership. When weighing structural options, the test is durable clarity of ownership and the effect on your key leaders, not tidiness on a slide.
The failure this altitude guards against is no longer a snatched-back task; it is key-person risk. An organisation whose critical initiatives destabilise when one director is courted by a competitor, or whose strongest VP refuses to develop successors, is fragile by design — and the accountable party is the executive who let it become so. Talent, likewise, belongs to the organisation, not to any one leader: a director who hoards strong engineers, blocking internal moves to protect local output, is optimising their team at the organisation's expense, and it is the executive's job to say so and open the paths.
Strong judgment looks like: designing structures where ownership survives any single departure — including your own; making succession an explicit expectation of every senior role; treating internal mobility as an organisational asset; delegating genuinely large mandates to VPs while keeping accountability undiluted.
Common pitfalls: becoming the decision bottleneck of a hub-and-spoke org; allowing critical functions to depend on one irreplaceable person; reorganising without weighing the impact on the leaders whose scopes change; confusing "empowered leaders" with "no one is accountable."
Self-check. Associate: what three things should an explicit delegation make clear, and why is task-relevant maturity a better guide than seniority? Professional: a manager who reports to you is drowning — what distinguishes coaching them from quietly doing their job? Expert: how would you detect key-person risk in an org you just inherited, and what makes succession a structural property rather than a personal virtue?
1.2 Feedback in both directions
Feedback is the mechanism by which behaviour changes. Most feedback fails not because it is wrong but because of how and when it is delivered: too vague to act on, too late to matter, or so cushioned the message disappears.
The most reliable structure separates observation from interpretation. The SBI model — Situation, Behaviour, Impact — anchors feedback in a specific moment ("in yesterday's review"), describes observable behaviour rather than inferred character ("you interrupted twice"), and names the consequence ("she stopped contributing"). It avoids diagnosing intent or generalising ("you always…"), which is what triggers defensiveness. Feedback should be timely, specific, behaviour-anchored, and two-way — a leader who never asks for feedback signals that feedback flows only downward. Praise follows the same rules: "great job" is forgettable; naming the specific behaviour and its impact makes it repeatable.
One law governs the higher altitudes: the truth degrades as it travels upward. Signal is filtered, softened, and summarised at every level between you and reality. The more senior you are, the more deliberately you must engineer channels that carry unfiltered information — and the more your visible reaction to bad news determines whether you ever hear it again.
At Associate — feedback with your team
The first-line version is direct: small corrections given frequently so feedback stops feeling like a verdict; hard messages delivered plainly and respectfully; clarity treated as a kindness. Upward feedback is solicited and visibly acted on.
Strong judgment looks like: frequent, small, SBI-shaped corrections; directness without cruelty on hard messages; soliciting upward feedback and closing the loop.
Common pitfalls: saving feedback for the performance review; softening a hard message until it's unrecognisable; targeting the person ("you're careless") instead of the behaviour; giving feedback while visibly frustrated, so the emotion drowns the content.
At Professional — feedback on leadership, at one remove
A manager of managers gives feedback on how someone leads, not just what their team ships — and must gather the evidence to do it, because a manager's own account is the most filtered source available. The instruments are skip-level 1:1s (whose purpose is unfiltered signal on team health and issues the manager might miss — never a bypass channel for decisions), direct observation, peer input, and team-health indicators like attrition and engagement.
The defining scenario: engineers quietly report that a high-output manager takes credit for their work and is dismissive in private, while the manager's own reports glow. Neither reflex is right — confronting the manager with raw hearsay burns your sources and invites denial; dismissing it because it hasn't reached a "formal channel" abandons the team. Strong judgment verifies discreetly through skip-levels and observation, then coaches the manager with specific, evidence-based feedback. The same discipline applies to the well-liked manager whose delivery has slipped for two quarters: likability is not the bar — name the actual gaps, with evidence, and set a clear path.
Strong judgment looks like: giving managers SBI-shaped feedback about their leadership behaviours; triangulating signal before acting on it, and protecting the people who provided it; asking your managers for feedback on your leadership and acting on it visibly.
Common pitfalls: giving managers feedback only about output and never about how they lead; treating second-hand signal as either a verdict or noise; letting a manager's self-report stand in for team reality; relaying skip-level input so identifiably that people stop talking.
At Expert — feedback as a system
At executive altitude, unprompted truth has largely stopped arriving. The work is to institutionalise what lower altitudes do personally: skip-level programmes, engagement data someone actually reads, calibrated leadership reviews, and — hardest — norms that make dissent safe at the leadership table. A leadership team that is homogeneous in background and converges on decisions without real disagreement is not aligned; it is unchallenged, and the executive's job is to change both its composition over time and its decision process now.
Seniority does not exempt anyone from hard messages. A respected director who avoids the strategic thinking their level now requires, or a leader whose operational excellence masks a growing gap, is owed the same directness as a junior engineer — delivered privately, specifically, and with a genuine path forward. The same holds between leaders: when two VPs are locked in unproductive conflict that is cascading into friction between their organisations, the executive addresses both directly — names the impact, makes repairing the working relationship an explicit expectation of their roles, and does not settle for routing around the feud or quietly picking a winner. What an executive owes their leaders is honesty, context, and candour about their prospects; comfortable silence is a form of neglect.
Strong judgment looks like: building mechanisms that surface bad news early and rewarding the people who bring it; engineering real dissent into leadership decisions; delivering hard feedback to powerful, senior people without flinching or humiliating.
Common pitfalls: surrounding yourself with agreement and calling it alignment; softening messages in proportion to the recipient's seniority; punishing messengers and then wondering why surprises keep arriving; mistaking the absence of complaints for organisational health.
Self-check. Associate: rewrite "you need to communicate better" using SBI. Professional: three engineers tell your skip-level something damning about their manager — what do you do first, and why not confront the manager immediately? Expert: what mechanisms replace the unfiltered signal you no longer get in person, and how does your reaction to bad news shape whether they work?
1.3 Hiring and onboarding
Hiring is one of the highest-leverage decisions a manager makes, and the leverage compounds with altitude: a bad engineering hire costs a team, a bad manager appointment costs every person on that team, and a bad executive hire can cost an organisation years. The competency is structured, calibrated evaluation rather than gut feel. Structured interviews — consistent questions mapped to a defined rubric, scored independently before discussion — are markedly more predictive and more equitable than unstructured "vibe" conversations. Calibration matters as much as structure: interviewers must share a standard for what "meets the bar" means. Guard against the halo effect (one impressive trait colouring the whole assessment) and affinity bias (favouring people like yourself).
Onboarding is where a good hire is either activated or wasted, at every level: a clear 30/60/90 framing, an early real contribution, an explicit guide, and documented context. The more senior the hire, the more onboarding is about integration — norms, relationships, and how decisions really get made — and the more its absence is the actual cause of failure.
At Associate — hiring engineers
Define the bar before interviewing; score against a rubric; protect the standard even when a seat has been open too long — a bad hire costs far more than an empty seat. Design onboarding so the first real win comes fast: a first task small enough to ship quickly, a buddy, documented context instead of tribal knowledge.
Strong judgment looks like: defining the bar first; structured, calibrated interviews; holding the standard under hiring pressure; onboarding engineered for an early real contribution.
Common pitfalls: lowering the bar because the search is dragging; over-indexing on a single impressive signal; "onboarding" that is a week of reading with no real task; assuming a strong hire needs no support.
At Professional — making and selecting managers
The Professional version of this competency is deciding who leads. The sound reasons to move an engineer into management are a genuine desire to lead, demonstrated people judgment, and the team's trust — not seniority, tenure, or retention. Management as a reward creates managers who don't want the job; the strongest IC who has repeatedly said they don't want to manage should not be pressured into leading your critical new team — understand what they do want, and develop or find another leader. When you do appoint, choose on demonstrated judgment in adjacent conditions, not on who is loudest or most available; and treat the first months as onboarding into a genuinely new profession, with coaching and protected room to learn.
External manager hires fail disproportionately at integration, not capability. When one is not landing at three months — thin results, a skeptical team — intervene early and honestly: diagnose specifically what isn't working, name expectations, and support the correction, rather than waiting politely for six-month clarity or concluding instantly that the hire was a mistake.
Compensation judgment also lives here. Faced with an engineer holding a competing offer, coach your manager to decide on value and fairness — if the person is underpaid, fix it because it's wrong, not because they threatened to leave; retention bought under duress distorts equity and rarely lasts. When a new hire's package inverts a tenured manager's, acknowledge it honestly and fix genuine inequity through the compensation process, not with improvisation or denial.
Strong judgment looks like: promoting people into management for the right reasons and supporting the transition deliberately; selecting leads on evidence of judgment; intervening early and supportively on a mis-landing hire; making compensation decisions on fairness rather than leverage.
Common pitfalls: using management as a seniority prize or retention bribe; drafting a reluctant IC into leadership; leaving an external hire to sink or swim; reactive counter-offers that teach people threats work.
At Expert — building the leadership team
Executive hiring decisions trade in different currencies. The classic: a critical VP seat, a promising-but-unproven internal candidate, a proven external hire. There is no formula — the judgment weighs the internal candidate's trajectory and organisational knowledge against the external's proven scale and integration risk — but the discipline is to make the bet consciously and then resource it: an internal promotion gets coaching and air cover for the step up; an external hire gets deliberate integration. A technically strong senior hire from a very different culture whose directive style is clashing is usually an integration problem before it is a capability problem — explicit norms, sponsorship, and early, direct feedback come before conclusions.
Beyond individual hires, the Expert owns the pipeline: a healthy leadership pipeline means key roles have ready-or-nearly-ready successors and rising leaders get real stretch opportunities before they're needed. It also means a leadership team diverse in background and thinking — a decision-quality issue, not a branding one: homogeneous teams converge fast and wrongly. And when you inherit a leadership team of mixed quality, assess each person individually against the bar over your first months and act deliberately — neither a loyalty purge nor indefinite drift.
Strong judgment looks like: making internal-vs-external bets consciously and supporting whichever is chosen; treating senior onboarding as the true risk surface; building a pipeline so no role has a single point of failure; composing a leadership team for genuine diversity of thought.
Common pitfalls: defaulting to the "safe" external hire and under-supporting them anyway; concluding a struggling senior hire "isn't a fit" without having onboarded them; a pipeline that is one name deep; cloning yourself at the leadership table.
Self-check. Associate: why are structured interviews more predictive and more equitable? Professional: what are sound versus unsound reasons to make someone a manager, and what do you owe them in their first six months? Expert: when does an unproven internal candidate beat a proven external one for a VP role — and what does the choice commit you to either way?
1.4 Performance management
Performance management is setting clear expectations, supporting people to meet them, and following through with honest consequences when they don't. At every altitude the foundation is the same: people can only meet a bar they can see, problems are handled early, privately, and specifically, and when performance falls short the first diagnostic question is skill or will — they don't know how, or they aren't applying themselves — because the responses differ entirely. A formal improvement plan, if it comes to that, is a fair, time-boxed, well-supported chance to succeed, not paperwork for a decision already made. Avoiding the hard conversation is the signature failure at every level, and its cost is always paid by the rest of the team, who see the gap and quietly recalibrate their own standards.
What altitude adds is distance and stakes: the Professional manages performance through and of managers, where errors compound across whole teams; the Expert judges leaders whose results are entangled with context, politics, and the executive's own public bets.
At Associate — managing individual performance
Set expectations people can see; address gaps early; separate skill from will; be clear and humane at once. Recognise that protecting the team sometimes means a hard decision about one person — including the "brilliant jerk" whose output does not offset the damage they do.
Strong judgment looks like: addressing issues before they become crises; distinguishing skill from will and responding accordingly; clarity as kindness; following through.
Common pitfalls: hoping problems resolve themselves; vague feedback that never names the gap; endless coaching of a will problem; tolerating a brilliant jerk.
At Professional — performance management at one remove
A manager of managers does two jobs here. The first is quality-controlling their managers' performance management. When a manager wants to put an engineer on a formal plan and you find expectations were never explicit and feedback was sporadic, the answer is neither rubber-stamping (the manager owns their team) nor taking over: decline the plan, coach the manager to set clear expectations and document honest feedback first, then revisit. The engineer hasn't yet been given a fair chance to succeed — and the manager is the one whose skill gap you just found. A manager who gives all high-visibility work to one or two favourites has an equity problem you name and require them to fix: opportunity is part of compensation.
The second job is judging managers themselves, which requires better evidence than their own reporting. A manager's true effectiveness is read from sustained team health and sustained delivery and how results are achieved — over time, not from a single quarter. Context matters: a promising manager's genuinely bad quarter driven by circumstances outside their control is weighed by the quality of their decisions and their response, not the raw outcome. Calibration exists to make this fair across teams: shared standards, evidence over advocacy, and when two of your managers can't agree on a rating for someone who worked across both teams, a joint review of the evidence — not seniority, and not splitting the difference. And when a manager completes a formal improvement period without meeting the bar, follow through: the decision was defined when the plan was set, and unwinding it teaches everyone the bar is negotiable. Harder still, when headcount must shrink and the manager disagrees with the selection, hear their case in full — then own the decision with criteria that are consistent and defensible.
Strong judgment looks like: coaching managers to do performance management properly rather than doing it for them; judging managers on evidence, over time, in context; running calibration as a fairness mechanism, not a ranking ritual; following through at the end of an improvement period.
Common pitfalls: approving a PIP built on absent expectations; judging managers by their own status reports or by likability; punishing outcomes without reading context — or excusing patterns because of it; letting calibration decay into horse-trading.
At Expert — judging leaders
Three tensions define the Expert band. Results versus methods: a VP who delivers strong results by leading through fear — while capable leaders quietly exit beneath them — is a performance problem, full stop. The results do not offset the damage, because the org's future capacity is the thing being spent; address it with the same directness as a delivery failure. Loyalty versus scale: a long-tenured, widely respected director whom the organisation has outgrown deserves honesty, not quiet marginalisation — an explicit conversation, a role reshaped to their real strengths, or a transition with dignity. Tenure earns respect and candour; it does not earn permanent scope. Your own credibility versus the evidence: when a leader you publicly bet on is clearly failing, acting on the evidence is the credibility move — executives who double down to protect their reputation lose both the org and the reputation.
Calibrating leaders across very different organisations is the Expert's fairness discipline: normalise for inherited state, constraint, and difficulty — judge the delta a leader produced and the quality of their decisions, not the healthiness of the hand they were dealt.
Strong judgment looks like: holding leaders accountable for how results are achieved; giving loyal veterans honesty instead of quiet demotion; cutting a failing public bet on the evidence; calibrating across orgs by delta and decision quality.
Common pitfalls: tolerating a feared-but-effective leader until the pipeline is hollow; letting loyalty freeze the org chart; doubling down on a failing appointment to save face; rating leaders by the size or health of what they happened to inherit.
Self-check. Associate: why is skill-versus-will the first diagnostic question? Professional: a manager brings you a PIP for an engineer who was never given clear expectations — what do you do, and what does it tell you about the manager? Expert: a VP hits every number and leads through fear — what exactly is the performance problem, and what does delay cost?
1.5 One-on-ones and career development
The 1:1 is the highest-frequency, highest-leverage leadership tool at every altitude — a recurring, protected space that belongs primarily to the other person, not to status. Status can travel by other channels; the 1:1 is for what doesn't surface otherwise: blockers, frustrations, growth, early signs of disengagement. People rarely leave suddenly — withdrawal, cynicism, a stopped flow of ideas appear first, in 1:1s, for leaders paying attention. Career development, likewise, is universal: understand where someone wants to go and create the experiences — not just conversations — that get them there, with ladders or competency matrices making "what does the next level look like?" concrete and fair.
What changes with altitude is what grows in the room: engineers' skills at Associate, managers' judgment at Professional, and the organisation's leadership capacity itself at Expert.
At Associate — the report's meeting
Hold 1:1s consistently even when busy — cancelling them whenever things heat up signals exactly the wrong priority. Let the report set much of the agenda, listen more than you talk, and connect day-to-day work to where they're trying to go. Retention is largely won or lost here.
Strong judgment looks like: consistent, protected 1:1s that belong to the report; open questions and real listening; growth built from real experiences; noticing disengagement early.
Common pitfalls: cancelling 1:1s under pressure; hijacking them for status; discussing career growth only at promotion time; talking more than listening.
At Professional — developing managers
With managers, the 1:1 becomes a judgment debrief: walk through the hard calls they're facing or just made, probe the reasoning, and coach the thinking rather than issuing rulings. Managers develop primarily through progressively harder real situations with coaching — courses and shadowing supplement, they don't substitute. A development plan for a manager is the same tool as for an engineer: specific gaps, real support, honest review — not a euphemism for managing someone out.
Around the 1:1 sit the Professional's wider instruments. Skip-levels gather unfiltered team-health signal (and are explicitly not a decision bypass). A remote manager whose team feels disconnected and who resists structure gets coached toward deliberate norms and rituals — distributed teams don't cohere by accident. Promotion fairness across teams needs shared criteria, evidence, and calibration, or each team's ceiling is set by its manager's assertiveness. And ambition is fuel to channel, not a threat to suppress: a talented manager who openly wants your role is an asset — give them real growth toward it and succession-relevant scope — though politicking that corrodes the team gets named directly, as behaviour, like any other.
Strong judgment looks like: 1:1s with managers that develop judgment, not just track status; stretching managers with real situations and debriefing them; building fair promotion mechanics across teams; channelling ambition into succession.
Common pitfalls: running manager 1:1s as project reviews; developing managers by course catalogue; letting each team invent its own promotion bar; treating an ambitious manager as a rival — or ignoring corrosive politicking because the manager is talented.
At Expert — building the leadership layer
The Expert develops leaders as a system. Succession exists for every key role — including the executive's own — and readiness is judged by whether someone is already exercising the next level's judgment on real stakes, not by tenure or by how much they want it. The classic dilemma — two directors ready for one VP seat — has no painless answer: make the honest call, tell the other the truth about their path, invest in it genuinely, and accept you may lose them; manufacturing a title to avoid the conversation weakens the org and fools no one. A VP who is excellent but refuses to develop successors has made their org fragile — make succession an explicit, non-optional expectation of senior roles.
Retention at this level is played long: a high-potential leader who is repeatedly courted stays for scope, growth, and a relationship that predates the counter-offer — not for reactive bids. What an executive owes their direct leaders is honesty, context, air cover, and real investment in their growth; and rather than shielding them from all organisational politics — which leaves them naive and blindsided — the executive translates: enough political context to operate, minus the noise. The measure that the whole system works is quiet: decisions that used to need you get made well without you, and the leaders you developed are developing leaders of their own.
Strong judgment looks like: succession as an explicit expectation of every senior role; readiness judged on demonstrated next-level judgment; truth-telling to the runner-up who didn't get the seat; retention built years before the poaching call; giving leaders political context instead of political shelter.
Common pitfalls: leadership development as an HR programme rather than an executive responsibility; equating readiness with tenure or ambition; inventing titles to dodge hard succession calls; counter-bidding for leaders you neglected; "protecting" leaders into naivety.
Self-check. Associate: why should the 1:1 be primarily the report's meeting, and where do the earliest attrition signals appear? Professional: what does a 1:1 with a manager develop that a status meeting can't, and how do managers actually get better? Expert: what is the strongest signal a leader is ready for a bigger role — and what is the strongest signal your leadership development is working?
Key takeaways
- Delegation keeps its anatomy at every altitude — outcomes, constraints, decision rights, retained accountability — while its object grows from tasks to teams to organisational structure; the matching failure grows from snatched-back work to bypassed managers to key-person risk.
- Feedback stays SBI-shaped and two-way at every level; what altitude adds is engineering the channels — skip-levels, calibration, dissent norms — because unfiltered truth stops arriving on its own.
- Hiring judgment scales from engineers to managers to the leadership team: the bar is defined first, management is never a prize, and senior failures are usually integration failures.
- Performance management is expectations, early honesty, and follow-through — applied at one remove to managers (coach, don't take over) and to leaders judged on how as well as what, in context, regardless of your public bets.
- 1:1s and career development compound: the report's meeting at Associate, judgment debriefs at Professional, and at Expert a succession system whose success is measured by how little the org needs you.