पाठशाला Pathshala · दल Dal, The team · Lesson 30 · Scale
The board and the founder-CEO
A board can make a founder-CEO better or make every decision twice. How boards change by round, how to choose the independent who holds the swing vote, and a CEO evaluation that strengthens rather than second-guesses.
Pathshala, The Founder Library · 11 October 2026 · 7 min read

A founder-CEO answers to a board for the first time somewhere around the first institutional round, and for most it is an awkward relationship for years. Run well, the board is the only group that can tell the CEO the truth with no stake in the answer but the company’s. Run badly, it is a quarterly audience that relitigates every decision and, eventually, decides the CEO’s future without the CEO in the room.
This lesson covers what the law asks of a private company’s board, how boards change round by round, choosing the independent director, meetings that inform rather than ambush, the CEO evaluation, and what to do when the board disagrees. The [board meeting lesson](/library/board-meeting-that-works) has the meeting itself.
What the board is, in law and in practice
The Companies Act asks little of a private company’s board. Section 149 requires at least two directors and allows up to fifteen, more with a special resolution. The independent director rules apply to listed companies and to unlisted public companies above set sizes of capital, turnover or borrowing, not to private ones. The board must meet at least four times a year, which the meeting lesson covers. And under section 169 the shareholders, not the board, can remove a director by ordinary resolution before their term ends, with special notice and a chance for the director to be heard.
Everything else comes from contracts. The shareholders’ agreement and the articles decide who nominates which director, which decisions need an investor director’s consent, and what happens to a founder’s seat if they stop being CEO. The [shareholders’ agreement lesson](/library/shareholders-agreement-what-you-are-signing) covers those clauses. Note the distinction they create: a founder can lose the CEO role by a board decision and still sit on the board, or lose the seat by a shareholder vote and still hold their shares. Samer Hamadeh and Adam Dinow, writing in TechCrunch, suggest founders negotiate a board seat that does not depend on the CEO role for exactly this reason.
Composition, round by round
Most venture-backed boards follow a recognisable path. Nadya Malenko’s research summary for the European Corporate Governance Institute describes it: founders typically control the board before the second venture round; in the second and third rounds the typical board is two investors, two founders and one independent director with the tie-breaking vote; by the fourth round investors hold most of the seats. Hamadeh and Dinow describe the same steps from the other side: after a seed round, two founder seats and one for the lead investor; a new seat for each round’s lead; an independent seat commonly added after the second round; and an odd number of directors to avoid ties.

Fred Wilson recommends a different shape for a founder-CEO: five seats, the founder-CEO, two independents and two investors, or seven with three of each. On that board neither side controls; the independents decide. Set your own board in the figure and see who carries a split vote.
The lesson of the figure is that the independent director is the most important seat a founder negotiates after the second round. When founders and investors disagree, which happens over a sale, a down round or the CEO’s own future, the independent decides. A founder who lets the investors pick that person, or leaves the seat empty until a crisis, has given away the swing vote without noticing.
Choosing the independent director
An independent director represents neither the founders nor the investors. Brad Feld puts the test as being an independent thinker, not a proxy for the investor or the CEO. Look for someone who has run a company a stage or two ahead of yours, ideally as CEO, who knows your market, and whom both founders and investors would trust to break a tie. A founder’s friend fails the test as surely as an investor’s former partner.
Agree the person with the investor directors, so neither side can later call them partisan, and do it in a calm year. Hamadeh and Dinow put early-stage independent director equity at between 0.5 and 2 per cent, falling as the company grows. Write down what you expect of every director, as the CEO in Feld’s post did: be honest, read the board book in advance, attend the four meetings, respond to clearly marked requests, support the company in public and use the product. A four-year term, renewed by choice rather than by default, keeps the seat working.
Meetings that inform rather than ambush
Most friction between a founder-CEO and a board comes from surprise. A director who hears bad news first in a board meeting assumes it was hidden; a CEO who hears a director’s objection first in the meeting feels ambushed. The cure is the rhythm in the [board meeting lesson](/library/board-meeting-that-works): a call with each director before the meeting, a pre-read five days ahead, decisions rather than reports, and the [board deck](/library/board-deck-what-goes-in-what-stays-out) built around them. Bad news goes to every director by phone within a day, never saved for the meeting.
End every meeting with a closed session of directors without management, then one with the CEO alone. The board expectations Feld published describe meetings that end with an executive session where directors give the CEO feedback and the lead director shares final thoughts. Ask for it rather than tolerating it: a CEO who gets feedback every quarter is never surprised once a year.
Choose the independent before you need the swing vote, and ask for your own evaluation before the board imposes one.
The CEO evaluation, run on the CEO’s terms
Every board evaluates its CEO, formally or in corridors. Hamadeh and Dinow note that the board approves the hiring and firing of senior management and sets the CEO’s pay; it is doing that judgement whether or not anyone writes it down. A founder-CEO is better off with a written, annual process they designed than with an impression formed in closed sessions they do not attend.
Run it in five steps, once a year, after the annual plan is approved. At the start of the year the CEO writes down what the board should judge them on: the plan’s three to five numbers, the two or three big decisions and the state of the leadership team. At the end the CEO writes a two-page self-assessment against those. The independent director, or a lead director if there is no independent, collects each director’s written view and anonymous input from the CEO’s direct reports. The independent delivers the synthesis to the CEO in person, then the CEO responds to the full board with the three things they will change. The board evaluates itself in the same session: attendance, preparation and whether each director is helping. The CEO’s pay is decided separately, without the CEO in the room.
When the board disagrees with you
Disagreement is the board working. The CEO’s job is to make it useful. Separate decisions the board must make, such as a fundraise, a sale, the budget or a senior hire under the shareholders’ agreement, from decisions it is advising on, and say which is which in the pre-read. On advice, listen, decide and explain the decision in the next update. On board decisions, find the objection before the meeting, address it or adjust, and bring the decision when the votes are known.
When a director keeps second-guessing operational decisions, raise it privately with the independent first and the director second, using the written expectations. And when the disagreement is about the CEO, listen hardest: a board that has lost confidence rarely says so plainly until it is late. The annual evaluation and the quarterly closed session are where a founder hears it early enough to change. Keep a short written log of each significant disagreement: what the board advised, what the CEO decided and how it turned out. Reviewed once a year, it shows both sides where the board’s advice has been right, where the CEO’s judgement has been, and which arguments are worth having again.
The founder-CEO’s board calendar
Every quarter: a call with each director three weeks before the meeting, the pre-read five days before, the meeting with closed sessions, a written summary the next day. Every month: a short written update to all directors, with bad news first. Within a day of any material surprise: a call to every director. Once a year: the CEO evaluation and board self-assessment, the director expectations reread, and a look at composition. Before every round: decide the board you want after it, use the figure to see who will hold the swing vote, and negotiate the seats and the independent in the term sheet rather than after it.
Board rights depend on your articles and shareholders’ agreement. Nothing here is legal advice.
Sources
- Companies Act, 2013, section 149: Company to have Board of Directors (with rules), ca2013.com, checked 11 October 2026
- Companies Act, 2013, section 169: Removal of directors, ca2013.com, checked 11 October 2026
- Nadya Malenko, Board dynamics over the startup life cycle, ECGI blog, November 2023
- Samer Hamadeh and Adam Dinow, What you need to know about startup boards, TechCrunch, 5 November 2016
- Fred Wilson, The Perfect Board, AVC, December 2014
- Brad Feld, Expectations for Outside Board Members, 1 April 2014