Leading Leaders: The IC to Product Leader Shift

Leading leaders means your job stops being about the work and starts being about the people who do the work. You no longer solve product problems directly — you build the judgment of the managers who solve them — and every instinct that made you a great individual contributor, and later a great first-line manager, now needs deliberate unlearning.

Quick answer: Leading leaders is a distinct job, not a bigger version of the one you had. Your output is no longer decisions or specs — it's the quality of the managers reporting to you. The pacesetting instincts that got you promoted are the ones you now have to actively suppress in favor of coaching.

Most product leadership advice is written for the transition everyone talks about: individual contributor to first-line manager. Far less is written about the transition that follows it, quietly wrecking careers that looked unstoppable a year earlier. Becoming a manager of managers — leading leaders — is where product organizations lose otherwise excellent people, not because they stop being good at product, but because they never stop being good at the wrong thing.

Why Leading Leaders Is a Different Job, Not a Bigger One

Leading leaders differs from every promotion before it because your unit of production changes completely. Instead of shipping features or coaching individual PMs directly, you're now responsible for the quality of decisions happening two levels below you — work you will mostly never see happen in real time, and can only influence indirectly through the managers you've developed.

Michael Watkins, whose book The First 90 Days remains the standard reference for leadership transitions, argues that every promotion carries its own distinct type of difficulty rather than simply "more of the same difficulty." His STARS framework (Start-up, Turnaround, Accelerated growth, Realignment, Sustaining success) was built for diagnosing business situations, but the same logic applies to the person: a leader-of-leaders transition is a different situation type than a first-time-manager transition, and treating it the same way is the first mistake.

The table below makes the shift concrete by comparing what actually changes at each level.

Level	Unit of Production	Time Horizon for Feedback	What Actually Gets You Fired
Individual contributor PM	Specs, decisions, shipped features	Weeks to one quarter	Weak product judgment
Manager of individual PMs	Coached PMs, healthy team rituals	One quarter to a year	Doing your reports' jobs for them
Leader of managers (Director+)	Capable managers, a functioning management system	A year to several years	Managing your managers as if they were still ICs

Read across the bottom row and the trap is obvious: the exact behavior that made you a strong manager of individual contributors — jumping in, modeling the right answer, staying close to the work — becomes the specific failure mode at the next level. Our complete guide to product leadership covers the earlier transitions in more depth; this piece picks up where that one leaves off.

The Leadership Pipeline's Hardest Passages — and Why This One Ranks Second

Ram Charan, Stephen Drotter, and James Noel's Leadership Pipeline model remains the most cited framework for describing exactly this problem: leadership development is a series of discrete passages, each requiring a genuinely new set of skills, time applications, and values — not just "more" of the previous ones. Skipping a passage, or pretending it isn't a real transition, is where the pipeline breaks.

The original model lays out six passages. The first two are the ones product organizations get wrong most often:

Managing self to managing others — the classic first promotion, usually cited as the hardest because it's the most disorienting: your success now depends on people instead of your own hands.
Managing others to managing managers — this piece's subject, and arguably the second hardest because it's the one companies most often skip preparing anyone for.
Managing managers to functional leader.
Functional leader to business leader.
Business leader to group leader.
Group leader to enterprise leader.

Charan, Drotter, and Noel are explicit that passage two fails constantly because organizations assume that if someone can manage people, they can automatically manage people who manage people. The two jobs share a title pattern ("manager") but almost nothing else. A manager of managers has to learn to value managerial work exclusively, select and develop first-line managers, and let go of any lingering identity as the most technically capable person in the room — precisely the identity a strong former IC still clings to.

That last requirement is the one nobody warns you about, and it's where the next section starts.

The Pacesetter Trap: When Your Best Skill Becomes a Liability

The pacesetting instinct — setting the standard yourself and expecting everyone to keep pace — is exactly what got you promoted twice already, and it's exactly what will stall you at this level if you don't retire it. Daniel Goleman's research on six leadership styles found pacesetting produces the most corrosive effect on team climate of any style when it's overused, because it silently communicates "I don't trust you to get there on your own."

Goleman's original study, published as "Leadership That Gets Results" in Harvard Business Review, found that pacesetting and coercive were the two styles most likely to damage morale and flexibility when relied on as a default. Authoritative and coaching, by contrast, correlated most strongly with positive climate over time — the same six styles our piece on Goleman's six leadership styles in the room walks through in more depth.

Pacesetting isn't useless — it works briefly with a small team of highly skilled, highly motivated experts who need no direction. It fails badly the moment your job is developing managers rather than deploying experts.

Watch for these signals that pacesetting has followed you into a leading-leaders role:

Rewriting a manager's document instead of coaching them to fix it themselves.
Sitting in on a direct report's 1:1s "just to help," uninvited.
Answering a stakeholder's question that belonged to your manager, in front of them.
Feeling a low-grade anxiety whenever you're not personally close to the details.
Measuring your own week by what you produced instead of what your managers produced.

None of these behaviors look like failure in the moment. Each one looks like helpfulness, speed, or high standards — which is exactly why the trap is so easy to walk into and so hard to notice from inside it.

Delegate Outcomes, Not Tasks: The Operating Shift

Delegating outcomes instead of tasks means you and your manager agree on the destination and the guardrails, and they own the method, the sequencing, and the daily calls entirely. The moment you start specifying how — which document format, which stakeholder to talk to first, which metric to lead with — you've quietly reverted to managing the task instead of the outcome, and you've taken the judgment-building rep away from the person who needed it.

The two modes look similar from the outside — both involve giving someone work — but they train opposite muscles, as the comparison below shows.

Dimension	Task-Level Delegation	Outcome-Level Delegation
What you specify	The steps, the format, the order of operations	The destination, the deadline, and the guardrails only
What they own	Execution of your plan	The plan itself, and the decisions inside it
What you check	Whether steps were followed correctly	Whether the outcome landed, and why it did or didn't
Effect on their judgment	Stays dependent on you for the next similar call	Compounds — each cycle sharpens their own calibration
Effect on your time	Scales linearly with headcount; you're a bottleneck	Scales with the number of managers you can trust

Read the last row twice: task-level delegation is the reason capable leaders quietly become the bottleneck in their own organization, because their time scales one-to-one with the number of decisions routed through them. Outcome-level delegation is the only version of the job that scales past a handful of direct reports.

In practice, shifting from the left column to the right one is a small number of disciplined moves, repeated consistently:

Define the outcome and its success metric together, not as an assignment handed down.
Set explicit decision rights — what they can decide alone, what needs a heads-up, what needs your sign-off — so ambiguity doesn't default to "ask me first."
Tie check-ins to outcome milestones, not daily task status; a standing 1:1 is for judgment, not status theater.
Ask coaching questions instead of supplying answers — "what have you already ruled out?" does more for a manager's growth than the correct answer would.
Let small failures happen before they compound into large ones; a manager who never sees the consequences of a call never builds the calibration to make better ones.

That fourth step is the hardest to practice consistently, because it requires you to sit with an answer you already know and not say it. It's worth building your own discipline around judgment the way you'd want your managers to build theirs.

The habit of writing a decision down before you know the outcome, reviewing it later, and being honest about what you'd change, is described well in our piece on using a decision journal to calibrate judgment. The tool works identically whether you're the one deciding or the one coaching someone else's decision — and modeling it openly is itself a coaching move.

The Emotional Cost of Giving Up the Hands-On Work

Stepping back from hands-on mastery carries a real sense of loss that's rarely named out loud, and pretending it isn't there just makes it leak out sideways as micromanagement. You spent years getting genuinely excellent at running discovery interviews, mapping a customer's emotional highs and lows, and writing the spec yourself — and the promotion asks you to mostly stop doing any of it personally.

Watkins' transition research is candid that even a "positive" transition carries an adjustment period with real psychological cost, not just a logistics problem to solve. The frameworks you once ran with your own hands — structured jobs-to-be-done interviews to find the real hiring criteria, or a full customer journey map to find where emotion craters — don't disappear from your job. They just move from something you do to something you teach a manager to do well, then watch them do imperfectly at first.

That shift in identity — from "the person who is best at this" to "the person who develops people who are best at this" — is a genuine grief, even when it's unambiguously a promotion in scope and compensation. A few honest markers of it:

A quiet resentment when a manager's version of the work is "good enough" but not how you'd have done it.
Missing the satisfaction of finishing a concrete artifact yourself.
Wondering, privately, whether your own skills are quietly atrophying.

There's a harder edge to this too. Occasionally, despite real coaching investment, a manager reporting to you simply cannot make their own transition — they keep doing their reports' jobs for them, no matter how much support you give. When coaching has genuinely been tried and hasn't worked, the healthiest thing for the team is usually a direct, respectful conversation about fit, handled the way our guide to managing someone out with dignity describes — not a slow, resentful drift that damages everyone involved.

Noticing the Shift in Real Time

You cannot reliably self-diagnose a pacesetting relapse in the moment, because it disguises itself as diligence — which is exactly why an external mirror matters more at this level than it did before. Gallup's long-running manager research has repeatedly put a manager's share of the variance in team engagement at somewhere around 70 percent, a rough but directionally telling reminder that at this level, your leadership style is the product, not merely an input to it.

For example, it can surface a pattern where your written feedback consistently defaults to rewriting a manager's work rather than asking the question that would let them get there themselves. It's built as a mirror for a blind spot, not a verdict; what you do once you see the pattern is still entirely on you.

Key Takeaways

Leading leaders is a different job, not a bigger version of the one you had — the unit of production shifts from your work to the quality of your managers' decisions.
The Leadership Pipeline model, from Charan, Drotter, and Noel, identifies this as one of the two hardest passages precisely because organizations skip preparing anyone for it.
The pacesetter trap is the single most common failure mode: the instinct that got you promoted twice becomes corrosive the moment your job is developing managers instead of deploying experts.
Delegate outcomes and guardrails, not tasks and methods — specifying "how" quietly reclaims the judgment-building rep your manager needed.
The emotional loss of stepping back from hands-on mastery is real and predictable; naming it early prevents it from leaking out as micromanagement later.
Gallup's manager-engagement research is a blunt but useful reminder that at this level, your leadership style is the deliverable.
External tools — including Prodinja's Leadership Styles module — can help surface a pacesetting relapse you're too close to see yourself.

Frequently Asked Questions

What is the hardest transition in product management?

Most leadership-pipeline research points to the first one — individual contributor to first-line manager — as the hardest, because it's the most disorienting: success suddenly depends on other people instead of your own output. Leading leaders (manager to manager-of-managers) is commonly cited as the second hardest, and the one organizations prepare people for least.

How long does it take to become a good manager of managers?

Expect meaningfully longer than a first-time-manager adjustment. Watkins' transition research frames the first 90 days as the acclimation window, but Charan, Drotter, and Noel's Leadership Pipeline research suggests fully internalizing a new leadership passage — the values shift, not just the skills — often takes closer to a year of deliberate practice.

What's the actual difference between managing others and leading leaders?

Managing others means coaching individual contributors on their direct work; leading leaders means developing the managers who coach those contributors. That requires you to value purely managerial work, judge managerial talent in others, and stop being the most technically hands-on person in the room.

How do I stop micromanaging as a director of product?

Start by naming the specific behaviors — rewriting documents, sitting uninvited in 1:1s, answering questions that belong to your manager — and replace each with a coaching question instead of a corrected answer. Delegating the outcome and the guardrails, rather than the task and the method, removes the opening for micromanagement by design.

Is becoming a director of product a promotion, or does it just feel like a demotion?

It's unambiguously a promotion in scope, influence, and typically compensation, but it can genuinely feel like a demotion of identity for a while. You're trading visible personal mastery for invisible influence through other people's work, and that emotional dip is normal and well documented in leadership-transition research — not a sign you made the wrong move.