K
KnowMBAAdvisory
LeadershipIntermediate5 min read

Skip-Level Meetings

A skip-level meeting is a 1:1 between a manager and someone two or more levels below them in the org chart — bypassing the immediate manager. Done well, skip-levels surface problems that filtered communication will never reveal: bad managers, broken processes, customer issues invisible to executives, and high-potential talent the system isn't recognizing. Andy Grove (Intel), Bill Campbell (the 'Trillion Dollar Coach' to Apple/Google/Intuit), and current operators like Brian Chesky (Airbnb) institutionalize skip-levels as the single highest-leverage information channel for senior leaders. The reason: organizations naturally filter bad news upward — your direct reports literally cannot tell you what's broken about themselves. Only their reports can.

Also known asSkip-Level 1:1sSkip LevelsLevel-SkippingN+2 Meetings

The Trap

Most skip-level meetings are useless theater because they're set up wrong. Three failure modes: (1) The senior leader uses the meeting to 'check up on' the middle manager — turning the skip-level into a covert performance review and destroying psychological safety. (2) Status updates: 30 minutes of 'what are you working on' — information the senior leader could have read in any doc. (3) The middle manager is told the skip-level happened, leading them to coach their report on what to say. Done badly, skip-levels actively erode trust between you, your direct, and their report. The meeting must be explicitly framed as 'not a back channel about your boss' — and yet it absolutely is one of the channels for surfacing manager issues.

What to Do

Run skip-levels with this structure: (1) Cadence: 30 minutes, every 6-8 weeks per skip-level report. Quarterly minimum. (2) Setup script: 'This isn't a status update and isn't a review of your manager. I want your view of the company, what's working, what's broken, and what would help you do your best work.' (3) Three core questions: 'What's something most leaders here don't understand about your team's work?' / 'What's slowing you down that I could help unblock?' / 'If you ran the company, what's the first thing you'd change?' (4) Confidentiality rule: themes go to the middle manager, attribution does not. (5) Always close the loop — within 30 days, report back what you did with the input.

Formula

Skip-Level Coverage = (Skip-Level Meetings Held per Quarter) ÷ (Total Reports 2+ Levels Below)

In Practice

Bill Campbell, who coached the founders of Apple, Google, Amazon, and Intuit, ran skip-level meetings as a religion. At Intuit (where he was CEO), he held weekly 'office hours' open to any employee at any level. He would walk the floors and talk to engineers, customer service reps, and finance analysts directly. These conversations surfaced the customer pain that became QuickBooks Online's mobile-first redesign — a decision the executive team would never have made on their own data. Campbell's mantra: 'The information you need to lead is being collected by the people you'll never see in your office unless you go find them.'

Pro Tips

  • 01

    Brian Chesky (Airbnb) skip-levels with random ICs picked by lottery monthly. The randomization prevents managers from prepping their 'best' people for the meeting and ensures unfiltered signal. Steal this pattern — randomize over a quarter, not over a year.

  • 02

    Always end with: 'Is there anything I asked that you didn't answer fully because you weren't sure how it would land?' This single question routinely surfaces 30% more useful information than the rest of the meeting combined. People hold back; you have to explicitly invite the held-back content.

  • 03

    Track patterns across skip-levels in a private doc. If 4 different ICs across 3 different teams independently flag the same broken process, that's signal. If one IC complains about their manager but no one else on that team does, the IC may be the issue, not the manager. Aggregation reveals what individual conversations conceal.

Myth vs Reality

Myth

Skip-levels undermine middle managers

Reality

Done badly, yes. Done well, they MAKE middle managers better. The good middle managers welcome skip-levels because they know it surfaces issues they can fix. Only weak managers fear them — and that's actually useful diagnostic information about who to develop further.

Myth

Senior leaders are too busy for skip-levels

Reality

Skip-levels REPLACE other lower-value time, they don't add to it. Two hours a week of skip-levels with high-signal ICs gives you 10x the org-health information of the same two hours in management status meetings. The opportunity cost of NOT doing them is invisible — until your best engineer quits and tells you in the exit interview things you should have known months ago.

Try it

Run the numbers.

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

🧪

Knowledge Check

You're a VP. After a skip-level with an engineer, she tells you (privately) that her manager is taking credit for her work. What's the right next move?

Industry benchmarks

Is your number good?

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

Senior Leader Skip-Level Cadence

Tech companies with strong feedback culture (e.g., Airbnb, Stripe, Shopify).

High-Signal Org

Every IC seen 2-4x/year

Average

Random / opportunistic

Insulated Leader

Never (filtered org)

Source: https://review.firstround.com/the-keith-rabois-method-of-coaching-the-next-generation-of-startup-stars/

Real-world cases

Companies that lived this.

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

💻

Intel (Andy Grove era)

1987-1998

success

Andy Grove institutionalized skip-level 1:1s as a non-negotiable management practice in his book 'High Output Management.' His rationale was direct: 'The single most important thing a manager can do is to know what's actually happening — and you cannot get that from your direct reports because they have a vested interest in what they tell you.' Grove ran skip-levels with engineers across Intel and credited those conversations with the early signal that pivoting from memory chips to microprocessors was urgent — a decision that saved the company in the late 1980s and made Intel's $300B+ run possible.

Intel Revenue (Grove tenure)

$2B → $26B

Pivot Decision (Memory → Microprocessors)

Informed by skip-level signal

Book That Codified Practice

High Output Management (1983)

Andy Grove's Cadence

Weekly skip-levels (rotating)

Skip-levels are not a 'nice to have' people practice — they are an executive intelligence channel that produces strategically decisive information. Grove treated them as central to running the company.

Source ↗
🏠

Airbnb (Brian Chesky)

2018-Present

success

Brian Chesky runs monthly skip-level lotteries: 4 randomly selected ICs from across Airbnb get a 30-minute slot with him. The randomization is intentional — managers can't prep 'their best people' for the meeting. Chesky has publicly credited these conversations with surfacing the critical signal that Airbnb's product was becoming too feature-bloated for its core hosts (leading to the 2022 'Airbnb 2.0' redesign focused on category browsing). The skip-levels also flagged morale issues during the COVID-era layoffs that the executive team hadn't fully appreciated.

Skip-Level Cadence

4 random ICs/month

Total ICs Per Year

~48 random samples

Product Decisions Influenced

2022 redesign cited publicly

Selection Method

Lottery (no manager filtering)

Randomization defeats the gaming problem inherent in skip-levels. Without it, you only meet the people your managers want you to meet — which is exactly the wrong selection bias for surfacing problems.

Source ↗
💥

Hypothetical: Series C Startup Skip-Level Disaster

2023

failure

Hypothetical: A 120-person Series C SaaS company's CEO started skip-levels after reading 'High Output Management.' Within his first round, he asked an engineer 'how's your manager doing?' and the engineer (cautiously) shared that their director was missing 1:1s. The CEO walked into the director's office that afternoon and said 'I heard from one of your reports you're missing 1:1s.' The director immediately knew which engineer it was. Two weeks later the engineer's project was deprioritized and his promotion was 'delayed.' Word spread. Subsequent skip-levels produced nothing but performative happy talk. Trust took 18 months to rebuild.

Useful Signal After Incident

Effectively zero

Trust Rebuild Time

18+ months

Engineer Outcome

Quit within 4 months

Root Cause

Confidentiality breach by CEO

Skip-levels die instantly the moment confidentiality breaks. One leak, one attributable consequence to a skip-level participant, and the entire information channel closes. THEMES go to the manager; ATTRIBUTION never does.

Related concepts

Keep connecting.

The concepts that orbit this one — each one sharpens the others.

Beyond the concept

Turn Skip-Level Meetings into a live operating decision.

Use this concept as the framing layer, then move into a diagnostic if it maps directly to a current bottleneck.

Typical response time: 24h · No retainer required

Turn Skip-Level Meetings into a live operating decision.

Use Skip-Level Meetings as the framing layer, then move into diagnostics or advisory if this maps directly to a current business bottleneck.