The board–CEO relationship is the one working relationship a chief executive can't escape and can't fully control: the board holds the ultimate authority, but the CEO holds the operating reality. Get it right and the board becomes the CEO's sharpest sounding board; get it wrong and every meeting becomes a defence hearing.

The quick version

  • The board's job is oversight and direction (hiring/firing the CEO, strategy, risk, integrity); the CEO's job is running the company. The whole relationship lives on that line, "nose in, fingers out."
  • Two theories explain the tension underneath: agency theory (the board polices a self-interested manager) and stewardship theory (the board partners with a trustworthy one). Real boards run on a blend, and the CEO's behaviour tilts which one applies.
  • Trust is the single biggest factor. A 2024 Spencer Stuart study of S&P 500 directors and CEOs found it ranks above almost everything else, and the fastest way to build it is a strict "no surprises" habit.
  • The best CEOs treat the board as a resource to recruit, not an audience to survive, they shape the agenda, brief the chair early, and let directors lead where the board genuinely should.

The idea in depth: two theories pulling in opposite directions

To manage the relationship you have to understand why it feels adversarial even when everyone means well. Governance scholarship offers two opposing accounts. Agency theory, set out by Michael Jensen and William Meckling in their 1976 paper on the firm as a "nexus of contracts," starts from a wary premise: owners (shareholders) hire a manager (the CEO) whose interests don't automatically align with theirs, so the board exists to monitor that manager and curb self-interested behaviour. On this view the board is a control mechanism, and independence, scrutiny and incentives are the tools.

Stewardship theory takes the opposite starting point. In their 1997 Academy of Management Review paper "Toward a Stewardship Theory of Management," James Davis, David Schoorman and Lex Donaldson argue that many executives are not opportunistic agents but stewards, collectivist, pro-organisational and trustworthy, motivated by doing the job well rather than by extracting from it. On this view, heavy monitoring is counter-productive: it signals distrust, demotivates a steward, and can produce the very behaviour it fears.

What this gives you is a way to read which mode your board is in, and to notice that you partly choose it. A board defaults to agency-mode monitoring when it feels uninformed, surprised, or unsure of the CEO's candour; it relaxes into stewardship-mode partnership when the CEO is consistently transparent, including about bad news. You can't lecture a board into trusting you. You can behave your way there. Every time you raise a problem before they discover it, the relationship moves one notch toward partnership.

flowchart TD
  A(["CEO behaviour over time"]) --> B{"Open about
bad news?"}
  B -->|"Yes, no surprises"| C(["Board relaxes into
stewardship mode"])
  B -->|"No, board gets surprised"| D(["Board tightens into
agency / monitoring mode"])
  C --> E(["More autonomy,
board as advisor"])
  D --> F(["More scrutiny,
board as auditor"])
Which mode a board adopts is partly the CEO's doing, candour earns autonomy. Leaders Loop

A caveat worth keeping. Neither theory is the whole truth, and the empirical record is mixed rather than settled. Studies comparing combined-versus-separate chair/CEO roles have produced conflicting results for decades, and most real directors hold both instincts at once. Treat agency and stewardship as two lenses for reading a given board's mood, not as a law that predicts behaviour. The durable insight underneath is simpler: a board's posture toward you tracks how safe it feels, and you influence that more than you'd guess.

The line: oversight versus operations

Most board–CEO friction is a boundary dispute. The board is responsible for direction and oversight, appointing and, if needed, removing the CEO, approving strategy, watching risk, and guarding the organisation's integrity, while the CEO and executive team are responsible for running the business day to day. The well-worn shorthand in governance circles is "nose in, fingers out": directors should be deep in the questions but out of the operational controls. As the Corporate Governance Institute puts it, a good non-executive director works like a sharp consultant, there to challenge thinking constructively, not to do management's job.

The two failure modes are mirror images. A board with its fingers in micromanages, undermining the executive and blurring accountability; a board that's too hands-off becomes passive, rubber-stamping and failing to hold management to account. Both are corrosive, and the CEO has real influence over which way a board drifts.

quadrantChart
  title Board engagement vs intrusion
  x-axis "Low intrusion" --> "High intrusion"
  y-axis "Low engagement" --> "High engagement"
  quadrant-1 "Micromanaging board"
  quadrant-2 "Effective: nose in, fingers out"
  quadrant-3 "Passive / rubber-stamp"
  quadrant-4 "Meddling but disengaged"
  "Healthy partnership": [0.28, 0.82]
  "Fingers in everything": [0.85, 0.80]
  "Asleep at the wheel": [0.20, 0.20]
The target is high engagement, low intrusion, a board deep in the questions but out of the controls. Leaders Loop

Better to make the line explicit than leave it to be litigated meeting by meeting. Agree, in writing if you can, what decisions are reserved to the board (the strategy, the budget envelope, major acquisitions, the CEO's own pay) and what is delegated to management. When a director starts reaching into operations, redirect rather than resist: "That's exactly the kind of execution detail my team owns, but the strategic question underneath it is a good one for the board." You're not fending the board off. You're inviting it back to its own altitude.

A board's posture toward you is largely a response to how safe it feels, and you control more of that than you think.

One honest caveat. Where exactly the line sits varies by company stage, ownership and jurisdiction, a founder-led startup board, a private-equity board and an ASX- or FTSE-listed board play very different games, and the legal duties of directors differ by country. Use the principle, but check your own governance framework and, on anything with legal weight, a qualified company secretary or adviser; nothing here is a substitute for that.

Trust beats reporting: the "no surprises" habit

If you could optimise only one thing in this relationship, the evidence points to trust. A 2024 Harvard Business Review piece, "5 Moments That Make or Break a CEO-Board Chair Relationship" (Anderson, Citrin, Frangos and Morfín), drew on a Spencer Stuart survey of nearly 200 directors and around 30 CEOs at S&P 500 companies and found trust to be the most critical factor separating effective relationships from broken ones. Notably, the authors found it isn't the volume of contact that builds trust, CEOs who exhaust themselves with pre-meeting calls to every director can actually undermine their own authority, but the timing and structure of communication, and above all, no surprises in the room.

The "no surprises" principle is exactly what it sounds like: a director should never learn something material about the business for the first time in the boardroom, in front of their peers, when they can't react well and feel ambushed. The fix is operational and unglamorous. Brief the chair before every meeting on anything contentious. The moment a number turns bad, a missed quarter, a key departure, a regulatory letter, the board hears it from you, early, framed with what you're doing about it, not discovered later from someone else. As governance practitioners put it, say what you'll do and do what you say; if you can't, the other party should be the first to know.

This is also why Boards That Lead (Charan, Carey and Useem, 2013) frames the relationship as a deliberate dance of "when to take charge, when to partner, and when to stay out of the way." Their argument, drawn from work with boards at firms such as Apple, Boeing and Ford, is that the best CEOs don't merely report to the board, they actively shape how it adds value, recruiting directors' expertise on the few questions where a board genuinely should lead, and protecting executive autonomy everywhere else.

A worked example

Consider a new CEO, call him Tom, six weeks into running a mid-sized software company, who has just discovered that a flagship customer renewal worth roughly a tenth of revenue is at serious risk. (Illustrative scenario; the people and figures are invented to show the method.) His instinct is to fix it quietly and mention it only if it falls over, so as not to look like he's losing control in his first quarter.

That instinct is the trap. If the renewal collapses and the board hears about it after the fact, worse, from the CFO or a market rumour, Tom hasn't just lost a customer; he's taught the board it can be surprised, which tips it straight into agency-mode monitoring for the rest of his tenure. Instead, he calls the chair the same week: here's the situation, here's the size of it, here's my recovery plan, and here's the one thing I might need from the board, an introduction to the customer's parent-company chair, whom a director happens to know. He's reframed a problem as a request for help.

At the board meeting, the renewal risk is on the agenda as a transparent item with a plan attached, not a confession dragged out under questioning. One director opens a door Tom couldn't. The deal is saved, but the more valuable outcome is durable: the board now reads Tom as a steward who brings bad news early, and it grants him more room as a result. The four minutes of discomfort in that phone call bought him months of autonomy, the exact inverse of what his first instinct predicted.

Frequently asked questions

What's the real difference between the board's job and the CEO's job?

The board owns oversight and direction: it hires and fires the CEO, approves strategy and major decisions, monitors risk and performance, and guards the organisation's integrity. The CEO owns execution: running the company day to day within the strategy and delegations the board has set. The shorthand is "nose in, fingers out", directors should be deep in the questions and out of the operational controls. Friction almost always traces back to one side crossing that line.

How often should a CEO talk to the board between meetings?

Regularly with the chair, sparingly with the full board, and always with a purpose. The 2024 HBR research found that more contact isn't automatically better, CEOs who try to pre-sell every director individually can look like they're managing the board rather than leading the company. A steady rhythm with the chair, plus prompt outreach the moment something material changes, beats a high volume of anxious check-ins.

My board keeps drifting into operational detail. How do I stop it without a fight?

Redirect to altitude rather than defend your turf. Acknowledge the underlying concern as legitimate, name it as an execution matter your team owns, and offer the board the strategic version of the question instead. If it's a recurring pattern, fix it structurally: a clear schedule of reserved versus delegated decisions, and a board agenda built around strategy and risk rather than operational status updates, gives directors a better place to put their energy.

Should the CEO also be the board chair?

It's one of the oldest debates in governance, and the evidence is genuinely mixed rather than decisive. Agency theory favours separating the roles so the board can independently oversee the CEO; stewardship theory is more relaxed about combining them where trust is high. Many jurisdictions and governance codes now lean toward separation or an independent "lead director" for listed companies, but practice varies by country and company stage, so check the code and norms that actually apply to you.

Isn't "managing the board" just political game-playing?

Only if you do it dishonestly. Managing the relationship well means being transparent early, shaping the agenda toward where the board genuinely adds value, and giving directors the context to govern well, the opposite of spin. The line is candour: a relationship built on no surprises and real information is strong precisely because it isn't a performance. If your standing with the board depends on them not knowing something, that's the warning sign, not the strategy.

Related in the Toolkit

Managing the board is stakeholder work at the highest stakes, so it builds directly on stakeholder mapping and on the discipline of managing up, a board is the most consequential "up" a CEO will ever manage.

Where to go next