What New CEOs Get Wrong About the Job

HBR IdeaCast (hosts Alison Beard, Adi Ignatius; guest Nitin Nohria) · podcast (HBR IdeaCast episode; transcript resolved from HBR's own unpaywalled website transcript, not Reader's broken 404 capture) · 2026-09-22 source ↗ Reader 🖥️

practical · argumentative — An interview (Alison Beard questioning Nitin Nohria) that is practical throughout: every segment builds toward specific advice for CEOs and CEO aspirants, grounded in Nohria's three decades running HBS's new-CEO workshop. It is argumentative rather than purely expository because Nohria repeatedly makes and defends specific claims (e.g., 'strategy is deciding what not to do,' 'humility predicts success') rather than just describing observations.

Structure

Key terms

• Running the company vs. running yourself out of the job
  The distinction between a CEO personally executing/deciding (as they did in prior roles) versus creating the organizational conditions under which a senior team executes and decides. Nohria treats direct CEO intervention as categorically different from the same act by any other executive, because of the outsized signal a CEO's involvement sends.
• Cornflakes / Frosted Cornflakes
  Nohria's borrowed metaphor for information distortion as it travels up to the CEO: raw facts ('Cornflakes') get progressively sweetened by subordinates managing their own image or protecting the CEO from bad news, arriving at the CEO's desk as flattering, softened reports ('Frosted Cornflakes').
• The board as your real boss
  A specific claim about the functional locus of accountability for a CEO: not any single person, but a 10-12-person collective body that is polite by norm, meets infrequently, knows less about daily operations than a conventional boss, and yet has sole hire/fire authority.
• Zero-based [team assembly]
  Borrowed from zero-based budgeting: a new CEO should evaluate their inherited leadership team as if assembling it from scratch today for their actual agenda, rather than defaulting to keeping people because they're already in the role.
• The 80/20 of the CEO job
  Nohria's claim that roughly 80% of what the CEO job demands (agenda-setting, communicating it, building a team, developing strategy, culture) has been constant across his three decades of observation, while about 20% is externally driven and changes by era (social-issue pressure, then geopolitics/tariffs, now AI).

Propositions

• explicitNew CEOs must deliberately stop running the company themselves and instead create conditions for their senior team to run it; direct CEO intervention reads to the organization as an outsized, unmistakable signal that disempowers subordinates and causes them to over-check every decision with the CEO, slowing the organization to a grind.conf: high
  supportsThe Best Leaders Embrace the Role of Supporting Character · p3 — Nohria's 'create conditions for others to run the company rather than running it yourself' is a direct restatement of 'the best leaders embrace the role of supporting character rather than main character,' applied specifically to the CEO transition.
• explicitCEOs (and their families) come under a categorically different level of scrutiny than any prior leadership role — every word and action is read for signal by insiders and outsiders, down to investors who hire specialists to analyze vocal inflection on earnings calls.conf: high
• explicitInformation systematically degrades in accuracy as it travels up to the CEO, because everyone reporting to a CEO has some self-interested or protective motive for presenting things favorably.conf: high
• explicitThe farther a source is from the people who report directly to the CEO — frontline employees, customers — the more likely their information is to be undistorted truth, yet time studies show CEOs spend the least calendar time with exactly those sources.conf: high
  extendsFrontline Workers Know How to Solve Your Organization's Biggest Problems · p10 — Both name the same structural gap (senior leaders are too distant from day-to-day reality to know what needs to change); Nohria adds the specific mechanism (information sweetens as it climbs the reporting chain) that explains why the frontline article's leaders are distant.
  supports7) Stine Haugseth: Insourcing, retning og strategi på en ølbrikke · p19 — Nohria's 'go to the front lines, talk to customers' advice is the English-language, CEO-specific instance of the same gemba/hospitering discipline Haugseth describes as how a toppleder stays epistemically grounded.
• explicitThe board, not any individual, is the new CEO's functional boss, and it is uniquely hard to manage: a 10-12-person body of formal equals, normatively polite rather than blunt, that meets only every two or three months and knows less about daily operations than a conventional boss — yet holds sole hire/fire authority.conf: high
• explicitEvery CEO Nohria has observed getting fired was surprised by it, because they mistook the board's habitual politeness for genuine support rather than privately probing individual members for unstated dissent.conf: high
  extendsIs Your Organizational Culture Too Nice? · p1 — Nohria's specific board-level finding (fired CEOs mistook the board's habitual politeness for real support) is a concrete instance of the general niceness-culture thesis that prioritizing harmony over candor compromises performance and accountability — here at the governance level rather than team level.
• explicitBecoming competent as CEO is primarily a learning-by-doing process with a steep early curve; most CEOs reach a working sense of mastery by their second or third year.conf: high
• explicitA reliable early red flag for likely CEO failure is the CEO's own leadership team being unable to accurately restate the CEO's stated agenda back, which Nohria tests directly by interviewing a random sample of a company's top 100 people.conf: high
• explicitPersistently dismissing analyst criticism ("the analysts don't understand my company") is a warning sign that a CEO is avoiding rather than engaging with legitimate investor concerns, and boards' patience for repeated bad quarters is finite.conf: high
• explicitHumility — remaining open-minded and continuing to learn — is the single trait Nohria most looks for as a predictor of a new CEO adjusting well, even though the job simultaneously requires projecting confidence to the organization.conf: high
• explicitA track record substitutes for a formal audition: people become credible CEO candidates chiefly by having, in one or two prior jobs, taken on a visibly risky assignment (a failing business, region, or new venture) and delivered on it.conf: high
• explicitEffective CEO time-management practice is a quarterly discipline: write down the agenda at the start of the quarter, then audit the actual calendar against it at quarter's end — if under half the time matches, the CEO's most scarce, non-fungible resource (their own time) is being misused.conf: high
• explicitAssembling a senior team fully aligned with the CEO's agenda is the CEO's single highest-leverage lever, and tolerating even one weak or misaligned member both distracts the CEO into over-managing that person and makes everyone whose 'shoulder is being watched' feel insecure.conf: high
• explicitEven in a stable, well-performing company, a new CEO should re-derive their leadership team from a zero-based starting point ("would I assemble this exact team today, for this agenda?") rather than defaulting to the inherited team; the most dangerous failure mode is knowing a change is needed and postponing it.conf: high
• explicitOrganizational motivation is sustained less by what a CEO says in the room than by "how do people feel when you're not in the room," which is driven by accumulated trust — earned through everyday, visible consistency between the CEO's actions, their stated strategy, and their stated values, including real consequences for people who violate the values.conf: high
  supportsHow Open Source Enables Collaboration in Creating a Platform · p3 — Both independently locate trust as an accretion built through everyday, consistent behavior rather than a one-time decision or a function of time elapsed — one about a CEO's relationship with their organization, the other about engineers' relationship with a platform.
• explicitOutsider CEOs are watched even more intensely than insiders at first, but visible early symbolic actions that respect the existing culture, combined with visible early change, let outsiders earn organizational trust within roughly six months — often faster than commonly assumed.conf: high
• explicitAbout 80% of the CEO job (agenda-setting, team-building, culture, strategy fundamentals) has stayed constant across Nohria's three decades of workshop observation; the remaining ~20% is generation-specific external pressure the CEO must respond to (social-issue activism in one era, geopolitics and tariffs in another, AI now).conf: high
• explicitCEOs are rarely blindsided by *what* is happening in the world; the harder, more common failure is knowing how to give a fast-moving external issue useful shape for the specific company — deciding what to actually do about it, not detecting that it exists.conf: high
• explicitStrategy is as much about deliberately deciding what *not* to do — including having the courage to say a domain isn't where the company will compete — as about deciding what to pursue; trying to win on every emerging front dilutes an organization's differentiated advantage.conf: high
  supports15) Kari Olrud Moen om strategi, kunsten å velge bort og forskjellen på å være tydelig og skråsikker · p1 — Near-identical convergent claim from an independent source: strategy is fundamentally the act of deselection — deciding what not to do — stated almost word-for-word by both Nohria (citing a colleague) and Olrud Moen.
• explicitNohria's top two pieces of advice to CEO aspirants: (1) don't look too far ahead — chronic future-focus visibly degrades present-job performance, so excel at the job you currently have; and (2) volunteer for opportunities others find too risky, since visible willingness to take on hard, unwanted assignments is what signals CEO-caliber capability to decision-makers.conf: high
  extendsThe Laws of this World (Game Theory) · p3 — Nohria's 'raise your hand for the risky assignment' advice is a specific application of costly-signaling theory: taking on visible, hard-to-fake risk is what makes competence for the CEO role legible to decision-makers who can't directly observe it.
• explicitThe CEO job carries a real, largely unavoidable personal cost — roughly a decade of consuming time demand, about a third of it traveling, with predictable trade-offs against family life — and aspirants should honestly assess whether they're ready for that cost before pursuing the role.conf: high
• implicitNohria's diagnostic framework locates the primary cause of CEO failure in the CEO's own behavioral and relational choices (poor board management, inherited-team inertia, insufficient humility, unclear agenda) rather than in structural or external constraints (industry headwinds, activist-investor pressure, macro shocks, weak predecessor conditions).conf: medium
• implicitThe framework treats "the truth about the organization" as a single, recoverable signal that already exists somewhere accessible (the front line) and merely needs a better channel to reach the CEO — it does not entertain that frontline perception is itself partial, locally biased, or that the organizational "truth" may be genuinely contested rather than merely filtered on its way up.conf: medium
• implicitNohria's board-management advice implicitly assumes a fairly uniform board behavior — "diplomatic politeness masking real dissent" — across very different governance contexts (public vs. private, founder-controlled vs. institutional-investor-controlled, activist-pressured vs. stable boards) without acknowledging that board incentives and power distribution vary by ownership structure.conf: medium

Arguments

Narratives

• A CEO's wife, walking through a grocery aisle, notices her own friends now looking at her differently after her husband became CEO.
  → Supports the proposition that CEO-level scrutiny extends beyond the CEO to their family, making the 'spotlight' collision more total and unexpected than new CEOs anticipate.
• Some investors reportedly hire people with intelligence-analysis backgrounds (described as 'CIA agents') to listen to CEOs' quarterly earnings calls for vocal tells, and an HBS colleague has written a case study on someone who tried to gain a trading edge this way.
  → Supports the proposition that external scrutiny of CEOs is extreme and highly specialized, not just casual public attention.
• A colleague from a cereal company gave Nohria the metaphor that 'information arrives to the CEO as Cornflakes and ends up on their desk as Frosted Cornflakes.'
  → Supplies the operative metaphor for the information-distortion mechanism (p3) — it does the explanatory work of the whole 'can't know everything' section rather than illustrating a separate point.
• Beard recounts hearing secondhand that Starbucks' new CEO (recently arrived from Chipotle) told staff to start talking to customers, and a barista repeated this instruction, unprompted, to a customer who asked what was going on.
  → Used by Beard (not Nohria) as a live, real-world example of an outsider CEO's symbolic early message visibly reaching and being internalized by frontline staff within a short window — illustrating both p16 (outsider trust achievable in ~6 months via symbolic action) and, obliquely, p4's front-line-truth thesis (the barista, not corporate comms, was the accurate, unfiltered source of 'what's going on').

Questions addressed

• ✅ What do executives (even experienced ones) get wrong about their expectations of the CEO role?
  Nohria gives a specific inventory: you can't run the company yourself, you're under total scrutiny including your family, you can't actually know everything despite having more access, and the board (not any individual) is your real boss.
• ✅ How can CEOs get more truthful information despite structural organizational filtering?
  Deliberately spend disproportionate time with frontline employees and customers, and cultivate a small circle of trusted, agenda-less 'truth tellers.'
• ✅ How do the best CEOs manage the board?
  Engage each member individually and privately, deliberately draw out quieter members, and never read collective politeness as evidence of actual support.
• ✅ Can you tell, before someone becomes CEO, whether they'll be good at it?
  Partially — look for a track record of at least one prior job where the person took a visible, real risk and delivered on it.
• ✅ What has changed about the CEO job over three decades, and what has stayed constant?
  Roughly 80% is constant (agenda, team-building, culture, strategy fundamentals); roughly 20% is era-specific external pressure the CEO must respond to (social-issue activism, then geopolitics/tariffs, now AI).
• ✅ Should a new CEO change the leadership team they inherited?
  Yes in principle — run a zero-based reassessment of whether you'd assemble this exact team today for your actual agenda, and don't postpone changes you already know are needed.
• ⭕ How should a new CEO distinguish appropriate hands-on intervention (in a genuine crisis) from the harmful 'running the company yourself' pattern Nohria warns against?

Gaps

• incompleteNohria never explains how the board-management advice ('engage members individually,' 'don't mistake politeness for support') should differ across governance structures — founder-controlled vs. institutional-investor boards, public vs. private companies, activist-pressured vs. stable boards. The advice is presented as if boards behave uniformly, but board incentives and power distribution plausibly vary sharply by ownership type.
  Read next: extract-strategi-som-regi-i-mote-med-teknologi-en-samtale-med-walter-qvam (Qvam on how ownership structure — state, dispersed retail, private equity, industrial families — shapes strategic discipline); general corporate-governance literature on activist-investor boards.
• incompleteThe rule 'don't run the company yourself' is stated as near-absolute, but real CEOs face moments (a safety incident, a fraud discovery, an active M&A crisis) where direct, hands-on intervention is clearly expected and appropriate. Nohria gives no criteria for distinguishing legitimate crisis intervention from the pathological over-involvement he warns against.
  Read next: Contingency-theory leadership literature on when centralized command is warranted vs. harmful (crisis management, high-reliability organizations).
• incompleteThe 'go to the front line for truth' advice treats frontline perception as a single recoverable signal, but doesn't address that frontline sources can themselves hold confidently contradictory reads on 'what's really going on' (a customer-service rep vs. a factory-floor engineer), and gives no method for reconciling or weighting conflicting frontline signals beyond 'talk to them.'
  Read next: extract-hbr-frontline-workers-solve-problems, which offers a structured method (interviews, customized surveys, cocreated/ranked solutions) for actually aggregating frontline input rather than relying on ad hoc conversations.
• incompleteThe claim that CEOs get fired because they 'mistook politeness for support' is Nohria's own retrospective pattern-read of anecdotal workshop conversations with already-fired CEOs. There's no base rate offered, and no comparison group — CEOs who did read the board accurately and were fired anyway, or CEOs who ignored this advice and survived — so the causal weight of 'board-reading skill' versus underlying company performance is left unclear.
  Read next: Empirical CEO-succession/dismissal research (e.g., studies correlating board engagement practices with tenure and forced-turnover rates) that isolates board-reading skill from performance as a predictor.
• incompleteThe claim that '80% of the CEO job hasn't changed in 30 years, 20% is external' is asserted from Nohria's own workshop observation, with no operational definition of what's actually being measured — time allocation, topics discussed in the workshop, or skill demands are three different things that could each yield a different ratio. As stated, the 80/20 split is a memorable number but not a falsifiable claim.
  Read next: None specific — this is a call for the underlying workshop data (if it exists) to be operationalized and published rather than asserted as a round number.

Closes gaps elsewhere

• 15) Kari Olrud Moen om strategi, kunsten å velge bort og forskjellen på å være tydelig og skråsikker · g3 — That gap asks how a leader earns and updates a credible read of 'how the world is moving' (forankring) rather than treating it as given. Nohria supplies a concrete mechanism: deliberately overweight time with frontline employees and customers over the normal reporting chain, and cultivate a small circle of agenda-less truth-tellers — because information demonstrably degrades in fidelity ('Cornflakes to Frosted Cornflakes') the closer it gets to the top.

Related documents

• The Laws of this World (Game Theory)
• How Open Source Enables Collaboration in Creating a Platform
• Is Your Organizational Culture Too Nice?
• The Best Leaders Embrace the Role of Supporting Character
• Frontline Workers Know How to Solve Your Organization's Biggest Problems
• Anders Aukland's Post
  Niceness Culture (Comfort Over Candor) Trust
• A Leading Restaurateur on What It Takes to Scale a Winning Culture
  Leader Restraint Trust
• When Persuasion Is (and Isn't) Manipulation
  Leader Restraint Trust
• How to Build a Superteam That Keeps Getting Better
  Leader Restraint Trust
• 7) Stine Haugseth: Insourcing, retning og strategi på en ølbrikke
  Go to the Floor (Gemba) Strategy & Direction-Setting