What is a board of directors for?

This piece began as an email to a CEO of a climate tech company I work with. He’d just raised a funding round and was putting together a board of directors for the first time. I tweaked that email and sent it to a few of my consulting clients in the ensuing months. It turns out lots of entrepreneurs struggle to conceive of, create and work with a board.

In October 2023, I attended Monktoberfest. It’s unquestionably the industry’s top technology conference with a craft beer focus. I pitched my board email as a talk and got accepted. That forced me to revise and expand my email. This is the result. I gave the talk from this script.

If you'd rather watch the talk than read it, you can do so on YouTube.

Entrepreneurs agonize over choosing co-founders. Who do you like and trust well enough to go through the category-five hurricane of creating a company together? They struggle with decisions about key hires. Which engineers are great enough? Which salespeople understand the product and market and know the customers? Which marketers can tell that story?

By comparison, almost every founder I talk to is surprisingly passive about their board of directors. A partner from the VC firm gets assigned to you. You have to take them, right? And they hardly ever think strategically about board composition or about how they engage with their board.

I think that’s because most entrepreneurs don’t know what a board of directors is for.

So I’ll explain that.

By the by-laws

When you incorporate your brand new startup, your law firm is going to send you a stack of documents. One of those is the by-laws, the rules by which you promise to operate. (You can and should tune those up, but that’s a topic for a different post). The by-laws capture some of the requirements of corporate law for the place you do business. Read them over. You'll see that the company promises to have a board of directors, and the directors promise to do some important work.

That work includes:

This is serious stuff. When your company is just a couple of people huddled around a kitchen table, you can kind of ignore it, but as soon as you start to grow, and especially if you take any money from investors, you need to get serious about it. You need board members who understand their responsibilities and can carry them out.

You should purchase a Directors and Officers (D&O) insurance policy no later than the time that you have a real board. This policy provides money to defend you and your board in the event of lawsuits. Company executives and board members are legally exposed, individually, to claims that they failed in their duties. Obviously they should make sure they don’t fail in their duties, but the company should also cover their legal defense to spare them out-of-pocket expenses if a lawsuit ever comes up. Nobody competent will agree to join your board unless you have a D&O policy.

Day to day

The by-laws are important, but they mostly live in a drawer. Day to day, you’re going to be focused on running the business, building product, finding customers, closing deals, finding partners, fending off competitors.

While you’re doing all of that, you’re going to come up against really hard problems. You’re going to need smart people you trust who can talk over the options, come up with ideas, make plans. You need to hire people who you want in the room for those discussions. They should make you smarter when stuff goes sideways. That’s your leadership team. They almost always have specific operational areas of responsibility and manage a chunk of the company. If someone isn’t on that list, then they’re not leadership.

Your board of directors should be an adjunct to your leadership team. The folks on payroll are in the business every day and know it best. Your board members should have insight and experience you can draw on to help you make tough decisions.

A good board member:

Who belongs on a board?

The company’s CEO has a voting board seat.

A board will also have a chairperson. Good governance suggests that the CEO and the chairperson are different people. When you’re tiny and have a small board of directors, you can fudge that, but as the board and the business grow it’s important to separate those responsibilities.

I strongly recommend, especially for the first few years of any company, that there be a second “common” (as in, holder of common stock) seat on the board. When investors buy stock, it’s generally “preferred.” The second common seat is often a co-founder, sometimes a trusted member of the leadership team. It helps you to have a second reliable vote for key decisions, though any split board vote that wins or loses by a narrow margin is a sign of dysfunction that needs fixing. But a vote you can count on is helpful.

You can move the common seat among folks on your leadership team over time.

So: two seats to the folks who run the company.

If you decide to raise money from investors, they should insist on board representation. That’s not always true for angel investors, but if you take a lot of money from someone and they don’t want board representation, it’s a worrisome sign. You’ll likely add someone from the venture firm or the angel group that gave you money as a voting board member.

When Amr, Jeff, Christophe and I decided to start Cloudera together, we knew we were going to need to raise money to hire a team. Ping Li at Accel Partners had convinced himself that the Apache Hadoop project was important. He was actively looking to fund a business to bring Google-style data analysis to the world. Nobody else we talked to was as clear about the opportunity. That made Accel’s money, and Ping’s seat on the board, an easy decision for us.

Not long after we got started, we received an unsolicited funding offer from a different venture firm. That was flattering! We didn’t need the money right then, but we knew we would in the year or so after, so we decided to take it seriously. With Ping’s help, we contacted a handful of other investment firms to see if they were interested. We wanted a market price for the business, and the only way to get that was to talk to several potential buyers.

We had lots of good conversations, but I was blown away by Aneel Bhusri from Greylock. The firm was made up of investors who had all been operators somewhere before. They’d had experience running real companies as C-level executives. Aneel had run Peoplesoft with Dave Duffield and was co-CEO at Workday. He was a serious operator. He was super smart and engaged. Like Ping, he’d developed an investment thesis that data was going to be important. Of course the economics of an investment matter – are you getting the best price? – but I thought that Aneel was valuable his own right. I wanted him on the board. It turned out that we got the right terms and I got the guy I wanted on my board of directors. 

Even if you don’t raise money, you’ll want useful board members. At Sleepycat, we added two independent members – not employees, not investors – to our board of directors. Guy Henshaw and I had served together on a biotech company board together. He was deep on finance and super helpful to the CEO during a successful sale of the business. Clyde Johnston had been the CEO of our largest customer, Innosoft, before they got bought by Sun Microsystems. I’d had lots of great conversations with him when he was my customer. When Sun bought them out, I wanted to keep him and his excellent business sense close.

I began the section by asking, who belongs on a board? As you can see, I think the answer is, some founders and some useful people.

That leads to the next question.

How can you recognize useful people?

Here are some useful questions to ask when you’re thinking about who to add to your board.

First, has the person ever held an operating role as a C-level executive in a company with characteristics like yours? Someone who's had to make payroll, plan layoffs, negotiate partnership deals, build a sales organization, open an office overseas, and so on is enormously valuable. You’ll likely have to do many of those things yourself in the coming years. The advice and support you get from another operator are hugely valuable to any CEO.

In its early years, Cloudera was a rocket ship. The market we helped to create was exploding all around us. We had lots of commercial opportunity. I had internal problems to deal with – founder drama, growth, sales execution, so much more.

I had a crazy Bay Area commute, ninety minutes in the car home at rush hour. Aneel Bhusri had a similarly unreasonable drive home. So many times, I called him from the car on the way home to talk through whatever ball of hair was most obstructive just at the moment. He’d challenge my thinking, talk through alternatives, encourage me. To the extent I got better as a CEO and Cloudera got better as a company in those years, much credit goes to Aneel and the time he spent with me. I’d been bullish on him as a board member before we had any real relationship.

Turns out I was right.

The second question to ask when assessing potential board members is, what connections does the person have that are useful to you? A former operator should have an address book full of contacts that will help you reach key potential partners and customers. They should help you find good candidates for key hires. And so on. A retired guy like me will have an address book that gets stale quickly, so it matters how recently they've been in the mix.

A partner at a venture capital firm has access to a firm-wide portfolio of companies. They should be able to put you in touch with other CEOs and leadership teams that are building growing businesses. That can be helpful. Those are people you can talk to about tough problems in your company who can give you advice and support. 

Greylock used to run a series of CIO bus tours. They’d bring a bunch of big-company executives to Sand Hill Road and ply them with fleece vests, soda and sandwiches. They’d invite portfolio company CEOs to brief them on the products we were building. I got to pitch Big Data to a captive audience of motivated buyers. This was a huge source of sales leads for us.

We also took some funding from In-Q-Tel. They ran an annual conference in San Jose that brought a bunch of mission-focused folks from the US intelligence and defense communities to a conference hotel downtown. In-Q-Tel helped me to book one-on-one meetings with groups already using the open source technology we’d built the company on, and those who had tough data analysis problems who could use our help. Working in those parts of government means you need to have staff with security clearances. I was a West Coast guy, mostly innocent of how any of that stuff worked. In-Q-Tel helped me understand the ecosystem and set up the processes and hire the staff needed to service that market.

These days, lots of venture firms have “platform teams” that are designed to assist portfolio companies with back-office stuff. It’s very common for them to offer some recruiting expertise. They may offer some finance, HR or marketing help. I think it’s nice if a firm offers you those services, but you should talk to the people on that team and make sure you understand what they do, and they do what you need.

Especially when you bring on independent members, contacts can be critical. Maybe you want an experienced marketer or someone with a legal background, because you need those skills in your boardroom. But someone who also knows your industry, has personal contacts with potential customers and partners, can really make a difference. 

I remember one board meeting at Cloudera where we were discussing partnerships. We realized it would be handy to get in touch with the CEO of a particular Fortune 50 firm. One of our board members piped up, “Hey, I’ve got her mobile number.”

That was nice!

The third question you should ask about a potential board member is critical: Can they help? You want to be in room with smart people who have your back and will plow their time, smarts and energy into your problems when they come up. This is really important! Board work is mostly easy and boring, but periodically spikes into brief periods of overwhelming intensity. Your board members should have the personality and the capacity to drop other stuff and help you when it's time.

In 2006, we sold Sleepycat to Oracle. Getting to that decision was hard. We had a good business and we liked running it. We were under-capitalized for the expansion we thought we needed to do. We struggled with a strategic decision: rais money, or find a strategic acquirer of the company who could make the investment we couldn’t? Clyde Johnston helped me think through the alternatives. Once we decided we wanted to find a buyer, he helped a lot with the bankers.

After we signed the Oracle term sheet, the diligence process kicked off. I’d never seen such a thing before and had no idea how to do it. Oracle was like fifty thousand people, and Sleepycat was twenty-five. They could roll over by accident and kill us.

The Oracle corp dev team quickly realized that I was out of my water. I didn’t know how to conduct a diligence process. They started to freak out a little bit.

My board member Guy Henshaw noticed my incompetence or their panic. One day, he just showed up at the office to help me run due diligence. He stuck around until we were done. 

I said above that you want smart people who have your back. That’s true. I think you also want people you like and trust.

Remember that the one real authority a board holds is to hire and fire the CEO. They might one day decide to move you out of your job. How sure are you that they are smart enough to approach that difficult decision with wisdom?

Finally – not least in importance, just last by chance on this list – you should think about specific skills that you need on the board. Are you doing business in a regulated industry? Maybe you need someone with a legal or regulatory background who understands the rules. Addressing weaknesses on the leadership team is important. If you’re a first-time CEO, or have a relatively inexperienced sales leader or finance person, you might want to put those skills on your board rather than hiring further up the food chain right away.

Board compensation

You and the other common seat don’t get paid extra for board service. Your bet is on the business. You have all the motivation you need to do a good job on the board.

Investors don’t need separate compensation either. They take their roles as stewards of the money they put in. They get paid when dividends go out or a liquidity event happens.

Independent members don’t have the founder or investor stakes, and you need to compensate them for the work they do. The usual thing when you’re still pretty broke is to grant them stock options and cover their travel expenses. They’ll often file an 83(b) election and exercise those options early. That’s a good vote of confidence in your business and has some real long-term tax benefits for them.

If your board chair is one of your independent members, they should get a bump in compensation for that role -- it's extra work.

If you’re profitable, you can pay your independent directors actual cash money.

Board terms

The CEO seat is essentially perpetual – it might be papered as fixed term with renewals, but you’re on the board. The other common seat might not live forever. A one-year or two-year renewable term makes sense. Your investors have a seat so long as their ownership stake warrants it. Your independent members should have renewable terms of two years or so, and you’ll need to renew their compensation when you renew their terms.

Rarely, but sometimes, the partner from a venture firm who sits on your board will leave that firm. The venture firm is the actual investor, so in that case, a different partner will usually assume the seat. You’re entitled to a board member that you want. You should vet the replacement using all the criteria I listed above.

Decisions that boards make

The board's role and responsibilities are primarily strategic. You know your mission. What strategy should you pursue in order to achieve it? The board should help you make that decision.

Here are some strategic questions that you might bring up in a board meeting.

Any questions on fundraising, share repurchase, new option pool allocations and so on must go to the board. The members have a fiduciary obligation to all shareholders and so must be consulted on this stuff.

In addition, any time you are considering hiring or firing a C-level executive, you want your board's endorsement ahead of time. You should have them interview key executive candidates as well. The decision on whom to hire is yours, of course, but if you bring on a COO who can't get along with your board, your life will be awful. Besides that, C-level execs have discretion that could lead to claims under a D&O policy if they screw up. The board is at risk in those claims so deserves to approve or reject those candidates.

You'll want the board to review and approve your pricing and discounting, the comp plan for sales, your employee leveling plans, your equity and compensation budget for staff, your annual budget and your hiring plans at a high level. You have the authority to hire line staff in the business, and to shift dollars among departments as needed, provided you're attentive to the overall budget. The board shouldn't be looking at staffing details lower down in the company.

As you can see, all of these are big, business-wide questions. The answers will generally be about major initiatives, with a high-level plan for how to pursue them, but with the detailed planning left to you and your leadership team.

By contrast, here are some decisions that aren't strategic:

A strategic question asks, "What do we need to accomplish to achieve our mission?" An operational question asks, "How do we do that?" The line between "where to go" and "how to get there" is blurry, and the board may have good ideas and advice on "how to get there." But at the end of the day, "how to get there" is a management decision.

You should be very, very careful about introducing operational questions in board meetings. You've got a roomful of smart people. They will absolutely discuss any question you bring to them. They will give you something between advice and orders. You ignore their advice at some peril, but you can't ignore their orders without putting your job at risk. You only want to ask questions where you need their advice or want their orders. And you should push back on them if they begin to give orders when they should only give advice. It's perfectly okay to say, "Okay, we understand the strategy, we appreciate the advice. We'll put a plan together and update you as needed on our progress."

Also, in general, you should assume that your board forgets almost everything between meetings. Sure, your board is probably better than that, but it’s best to be safe. Context is critical to quick and correct decisions. So when you bring an issue to them, be sure to provide background -- what the situation is, any critical history, what they decided in prior conversations, why this is an important question for them to consider now.

Running a good board meeting

You'll want to update the board during your meetings on your operational progress against metrics that matter to the business.

It’s a good idea to bring in key operational leaders to brief on operations. They should prepare the material in advance and rehearse it with you before the meeting. You should coach them to make sure the discussion is board-relevant. They should have a time budget and they should stick to it. They should absolutely be present for and involved in strategic discussions that touch on their areas of responsibility. It’s good professional development for them to participate in board meetings, and it’s good for your board to know your leadership team.

The read-out portion of the meeting, where you report on operational metrics, is critical. It should also be as short as possible. If you’re spending most of the meeting here, you’re wasting your time and your board members’.

The bulk of every board meeting should be dedicated to a key strategic question or two. Your job as CEO is to decide what each meeting is for, and to craft the agenda and material to support the update and cover the strategic topics. You should walk out of the meeting with a clear understanding of the strategic decisions, and those should be memorialized in the meeting minutes.

You, your chief of staff or your CFO should assemble board materials and deliver them to the board a week in advance. You're entitled to a board that reviews your material carefully ahead of time and shows up prepared. Ideally you have a dashboard that highlights your points, and you talk to that; there is no doubt background material in the deck, but you don't walk through every slide of detail. You hit just the essential points. The board packet should include the formal resolutions on which you plan to vote. Your attorney can help you by drafting those, and your board members can help you figure out what the right resolutions for the meeting will be.

You must have a designated secretary to take minutes during the meeting. That has to be someone who knows what board minutes look like. If you don't have such a person right now, find out if one of your investors has an associate who could join the member at the meeting and serve that function. Else, your lawyer can serve in that role for hundreds of dollars per hour. You need to review the minutes immediately after the meeting for correctness and completeness. You should circulate them immediately to board members for review. You'll vote to approve the previous meeting minutes, and to approve any other resolutions requiring a formal board vote, at every board meeting, typically at the end (because you want the discussion on the topics to happen before the decisions).

Your board minutes and meeting material are critical corporate documents. You can sign up for a service like Boardable to manage them, or you can keep track of them on your company systems. Either way, you must be careful in preserving those records. They’re necessary in audits and in M&A activity, and also so you remember what you decided and what budget got approved.

Meeting cadence depends on your business. Early on at Sleepycat and Cloudera, we met monthly. Over time, we switched to quarterly meetings. We didn’t need key strategic decisions from the group more often than that (and you can always convene a special meeting if you do). Putting the material together, preparing your team, spending the time, documenting the outcome and circulating the results is time-consuming! You shouldn't do it more often than you need to.

I’m on one board now. We meet quarterly for several hours three times a year, with a separate all-day meeting once a year to review budget, financial plans, targets and so on for the coming fiscal year. That session also generally considers some really big strategic topic or two, meatier than what we do in our shorter meetings. We bring the board together for a dinner the night before and invite key people in the company to join.

Of course that’s a regular-course-of-business schedule. From time to time, exceptional circumstances will require that you call the board together off-schedule. Fundraising, acquisitions, lawsuits, major competitive or market problems are all reasons to do this. If you’ve chosen your board members well, you’ll want to talk to them in all those circumstances, because they’ll be helpful to you!

Walter Isaacson tells an interesting story about Apple’s board in his biography of Steve Jobs.

In 2010, the Apple 4 came out. It was a nice phone, but soon after release, people started noticing that when they held it in their hands, their carrier signal would drop to roughly zero, and they couldn’t make phone calls. This was a big deal, because back then, people still made phone calls.

Apple initially announced that the problem was because people were holding their phones wrong. Even at the time, that seemed like a stupid excuse. The company’s refusal to acknowledge the problem led to “Antennagate.” People figured out there was some kind of inductive interference from your skin that made the antenna ineffective, and Apple was taking our money and trying to play the problem off as idiot user error.

Isaacson reports that Jobs pulled his board together for a series of intense working sessions to deal with the problem. Part of the discussion was crisis communications: What should we say? Part of it was short-term problem-solving:  We’ve sold a bunch of these phones and would prefer not to recall them. Is there anything we can do? (They gave away free cases.) And part was addressing the real long-term issue: How do we fix the antenna? What do we do with our manufacturing processes and supply chain, all optimized for a flawed design? 

I said above that boards should be strategic, not operational. But in a crisis, you’ll want to get operational advice from your board. Steve Jobs certainly did in this instance. And, as Isaacson documents, Apple clawed its way out of Antennagate.

When things get hot for you, convene your board. Explain the problem so they know what’s going on. Get their help.

Every single board meeting should end in an executive session, followed by a closed session.

For the exec session, everyone who is not an actual board member -- all of your staff who reported on ops, for example, leaves. You and the other internal member stay. Obviously the secretary has to stay to take notes. The board discusses anything it likes without those people present.

Then you and your co-founder common seat holder leave. The remaining board members have time in the closed session to talk amongst themselves without you in the room. Immediately after the closed session, one of your board members should follow up with you directly -- phone, video or in person -- with any instructions or guidance for you and the team that came up in that session.

The intent of the exec session and closed session is to provide opportunities for candid discussion. You want a frank and honest board. Sometimes there won't be anything to talk about in the exec or closed session. That's fine. But if you do them all the time, then you don't create any anxiety by scheduling them when you do need them.

At least a week, but no more than two weeks, before each formal board meeting, you should spend a half hour or so with each of your board members. You'll want to walk through the planned meeting agenda and get their buy-in on the topics. If there's anything a member wants to add, you should accommodate -- but pay attention to my warnings on strategic versus operational content above! You report on operations. You consult on strategy.

It's a good idea to begin each board meeting with a brief executive session, just the board members, no more than fifteen minutes. You should provide a quick summary of who’s coming in and what the key take-aways will be. Coach them on issues that need their attention, questions worth asking or praise worth delivering. Getting that directly from board members is hugely impactful to staff. Then, open the doors and let all the presenters and your leadership team come in.

The agenda and materials for board meetings are confidential and should not be shared outside of the board. Your staff presenting will see it, of course, but must keep it confidential. You should not share it with others who don't have legitimate access under your governance rules.

I say above that you should have quarterly formal board meetings. Of course you'll talk to your board members more often than that. You should reach out to individuals whenever you need their advice or help, even on operational stuff. You should lean on them for introductions to candidates, prospects, partners, etc. They should be inviting you to events they hold for their portfolio company leaders.

You shouldn’t schedule regular standing meetings with your directors. You’re busy and so are they. Cluttering up calendars diminishes the value of the time you spend together. Talk to them when you need them, and let them pay attention to other things otherwise.