Founders tend to fill early board seats with people who are easy to work with: early investors, friends with relevant experience, people who won't push back too hard in a meeting. That's a comfortable board, and it's usually the wrong board for a company that's trying to scale past where the founder's personal network and instincts can carry it alone.
Recruit for the Gaps, Not the Relationships
I now think about board composition the way I think about executive hiring: what specific expertise does this company need in the next eighteen to thirty-six months that nobody currently in the room has? If the company is heading toward international expansion, I want someone on the board who's actually done that, not just someone I like having dinner with. If a sale process is on the horizon within a few years, I want board experience with M&A specifically, because that's a different skill than early-stage growth advising.
That means recruiting board members who will occasionally disagree with me, and building the governance maturity to actually want that friction rather than avoid it. A board that only ever agrees with the founder isn't providing oversight, it's providing validation, and validation doesn't catch the mistakes that actually sink companies.
Refresh It Deliberately, Not Reactively
I now treat board seats as term-limited in practice even when they're not contractually structured that way, revisiting composition annually and being honest about whether each director's expertise still matches what the company needs at its current stage. A board built for a five-person startup rarely fits a hundred-person company, and holding onto those original seats out of loyalty is a governance failure dressed up as gratitude.
The right board isn't the most comfortable one. It's the one that can actually see around corners the founder can't see alone.