LeadershipIntermediate5 min read

Coaching vs Mentoring

Coaching and mentoring are routinely conflated but require fundamentally different conversation modes. Coaching: the leader asks questions to help the person solve their own problem — the answer comes from THEM. Mentoring: the leader shares relevant experience and direct advice — the answer comes from the MENTOR'S past. Bill Campbell ('the Trillion Dollar Coach') famously said: 'I don't have advice. I have questions.' That's coaching. Reid Hoffman publicly mentoring early founders by sharing what worked at PayPal — that's mentoring. Both are valuable; neither substitutes for the other. The skill is diagnosing which mode the person needs in this moment, then committing to it. Mixing them creates the worst of both: vague questions followed by your unsolicited opinion.

Also known asCoachingMentoringCoach vs MentorLeadership Development Conversations

Challenge a friend Browse library

The Trap

Most managers default to mentoring (telling) when the person needs coaching (asking). Why: telling is faster, feels productive, and validates the manager's experience. The cost: the team member never develops their own decision-making muscle. They become dependent on the manager for answers, can't operate when the manager isn't there, and don't grow into leaders themselves. The opposite trap (less common but real): coaching when mentoring is needed. A first-week new hire asking 'how do I submit my expense report' doesn't need Socratic questions — they need the answer. Pure-coaching-as-religion can become its own form of withholding. The senior leader who refuses to share their experience because 'they want them to learn for themselves' is sometimes hoarding insight while pretending to develop people.

What to Do

Diagnose mode at the start of every development conversation. Ask: 'Do you want me to think through this with you, or do you want me to share what I'd do?' Coaching is for when the person has the experience to find their own answer (they're stuck, not ignorant). Mentoring is for when they lack the context (they don't know what they don't know). Default ratio for managers: 70% coaching, 30% mentoring. For senior leaders coaching their direct reports (who are themselves senior): closer to 90% coaching. Use the GROW model for coaching conversations: Goal (what do you want?), Reality (what's the current situation?), Options (what could you do?), Will (what WILL you do, by when?).

Formula

Coaching Effectiveness = (Questions Asked × Time Person Spent Talking) ÷ (Statements You Made × Time You Spent Talking)

In Practice

Bill Campbell coached the founders of Apple, Google, Amazon, and Intuit for over 20 years — not by giving them strategic advice, but by asking questions that forced them to articulate what they actually believed. Eric Schmidt (Google's chairman) recounted that in critical moments, Campbell would ask 'what do you think is going on here?' instead of telling Schmidt what to do. The discipline of withholding advice was Campbell's superpower. Contrast with Reid Hoffman, who explicitly mentors founders by sharing what he learned at PayPal/LinkedIn — different mode, also enormously valuable. Neither pretends to be the other. Confusion happens when leaders try to be both simultaneously and end up doing neither well.

Pro Tips

01
Track your talk-to-listen ratio in 1:1s. In a true coaching conversation, the team member should talk 70-80% of the time. If you're talking more than 30%, you're mentoring (or worse, lecturing). Most managers radically overestimate how much they listen — record and time yourself for 2 weeks if you're skeptical.
02
The single most powerful coaching question is 'and what else?' After someone gives you their first answer to a question, ask 'and what else?' Their second answer is usually deeper than the first. Ask a third time. By the third 'and what else,' people often surface what they've been avoiding saying.
03
When mentoring, anchor advice with the specific context that produced it. 'When I was at Stripe, we faced X with Y constraints, and we did Z because A and B' is genuinely useful. 'You should do Z' is dangerous because the listener can't tell what to extract for their different situation. Untethered advice is malpractice.

Myth vs Reality

Myth

“Senior leaders should always coach, not mentor”

Reality

Bad people-development advice. Mentoring (sharing experience) is critical for less-experienced team members who lack the context to coach themselves to a good answer. Pure-coaching dogma withholds value. The right ratio depends on the gap between mentor's expertise and mentee's experience — wider gap, more mentoring; narrower gap, more coaching.

Myth

“Coaching takes longer than just giving the answer”

Reality

True for the single conversation, false over time. Mentoring creates dependency — you'll be giving the same person the same kind of advice 20 times over a year. Coaching creates capability — that person learns to find their own answer and stops needing you. Cumulative time: coaching wins by 5-10x within 12 months.

Try it

Run the numbers.

Pressure-test the concept against your own knowledge — answer the challenge or try the live scenario.

🧪

Scenario Challenge

Your direct report (a senior PM with 5 years of experience) comes to you upset about a conflict with the engineering lead over prioritization. She says 'I don't know what to do — should I escalate to the CTO?' You can either coach or mentor.

Industry benchmarks

Is your number good?

Calibrate against real-world tiers. Use these ranges as targets — not absolutes.

Manager Talk Ratio in 1:1s (% of time manager is talking)

Self-recorded data; most managers underestimate by 15-20 percentage points.

Coaching Discipline

20-35%

Mixed

35-60%

Lecture Mode

60%+

Source: International Coaching Federation research

Real-world cases

Companies that lived this.

Verified narratives with the numbers that prove (or break) the concept.

🏈

Bill Campbell ('Trillion Dollar Coach')

1980s-2016

success

Bill Campbell coached the founders of Apple (Steve Jobs), Google (Larry Page, Sergey Brin, Eric Schmidt), Amazon (Jeff Bezos), and Intuit (Scott Cook), among many others. The companies he coached are collectively worth over $2 trillion. His method was almost purely coaching — Schmidt, Rosenberg, and Eagle (his coachees who later wrote 'Trillion Dollar Coach') describe him asking questions instead of giving answers, even when he clearly had strong opinions. Campbell would say 'I'm just a football coach' (he had been a college football coach before tech) — meaning his job was to help the team find their own play, not to call the play himself. His refusal to take credit and to give direct advice was the SOURCE of his enormous influence.

Companies Coached (combined value)

$2T+

Years Coaching Tech Founders

30+

Public Speeches Given

Almost zero

Books Written by Campbell Himself

Zero

The most influential development relationships are usually coaching, not mentoring. Campbell shaped the most valuable companies in history by withholding his own answers and forcing leaders to develop their own.

Source ↗

🔗

Reid Hoffman Mentorship Network

2003-Present

success

Reid Hoffman (LinkedIn co-founder, Greylock partner) is widely regarded as one of Silicon Valley's most active and effective mentors. His method is explicitly mentoring (not coaching) — he shares specific patterns from PayPal, LinkedIn, and his portfolio of 100+ investments. He authored 'Blitzscaling,' 'The Alliance,' and 'The Start-up of You' to scale his mentoring through written content. Founders he has mentored include Mark Zuckerberg (Facebook), Drew Houston (Dropbox), and Kevin Systrom (Instagram). His mentoring is HIGH-CONTEXT — he gives specific, situated advice tied to specific patterns from his experience. This is mentoring done well: experience shared with anchoring context, not floating advice.

Companies Mentored / Invested In

100+

Books Written to Scale Mentoring

Mentees Who Later Founded Major Companies

Zuckerberg, Houston, Systrom, etc.

Mode

Explicit mentoring (not coaching)

Mentoring done well is enormously valuable — when the mentor has experience the mentee genuinely lacks AND when the advice is anchored to specific context. Hoffman models the upper bound of mentoring; Campbell models the upper bound of coaching. Both are correct in their lane.

Source ↗

🌀

Hypothetical: Mode-Confused Manager Failure

2023

failure

Hypothetical: A senior engineering manager at a 200-person company prided himself on 'always being available' to his team. His pattern: he'd ask coaching-style questions ('what do you think we should do?'), wait 5 seconds, then immediately tell them what HE thought they should do. His 7 reports learned to wait through the question to get the real answer. Within 18 months, his team's promotion rate was the lowest in engineering — none of his ICs could articulate decisions independently in promotion committee. Two of his strongest engineers transferred to other managers explicitly citing 'I want a manager who doesn't answer his own questions.' The manager genuinely believed he was coaching — he was actually performing coaching as a prelude to mentoring.

Promotion Rate (his team)

Lowest in engineering

Engineers Requesting Transfer

2 of 7

Self-Identified Mode

'Coaching'

Actual Mode

Performative coaching → directive mentoring

Mode-confusion is worse than honest mentoring. If you're going to tell people what to do, just tell them — don't first ask them what they think and then ignore the answer. Performative coaching destroys trust and produces no development.

Related concepts