July 29, 2026 · Investment Committee, Process
How to run an investment committee that decides
By Emily Buckley, founder of SamCIO
The failure mode of a bad investment committee is not conflict. It is a ninety-minute meeting that ends with "let's get more information," a phrase that has never once been a decision. Learning how to run an investment committee that actually decides is mostly about removing the structural excuses for not deciding, because the people in the room are rarely the problem.
Three things separate committees that decide from committees that convene: the reading happens before the meeting, the vote is on the record, and dissent survives contact with the outcome.
The memo circulates first, or the meeting is a reading group
If partners arrive having not read the memo, the first thirty minutes become a verbal summary. Verbal summaries are shaped by whoever is presenting, which means the meeting is now assessing a pitch rather than a document. Every discipline you built into the memo template is bypassed in the retelling.
The fix is unpopular and effective: circulate the memo a fixed number of hours before the meeting, and if it did not circulate, the deal does not get discussed. Not "we will do our best." The deal moves to the next meeting. Two enforcements and the norm holds permanently, because the cost of slipping is now visible and lands on the sponsor rather than on the committee.
This only works if writing the memo is cheap enough that the deadline is realistic. A committee that demands a twelve-page document per deal from a three-person fund will get late memos forever, and the rule will be the thing that breaks rather than the behavior. Consistency of structure beats length. A memo that hits the same sections every time is comparable across deals, and comparability is most of the value.
Ask for a decision, not a discussion
Discussion expands to fill the time available. The single most effective structural change is to open with the recommendation and the question being decided, rather than arriving at it after forty minutes of context.
"I am recommending we lead a $2M round at a $14M post. The question for the committee is whether the go-to-market risk is acceptable at that entry price." Now the discussion has a shape. Someone who thinks the price is fine but the market is wrong has to say so explicitly rather than gesturing at general unease for ten minutes.
The corollary: every deal on the agenda gets an outcome before the meeting ends. Advance, pass, or a specific conditional with a named owner and a date. "More information" is only a legitimate outcome when the committee can say what information, who is getting it, and what decision it would change. If nobody can answer those three, the committee is not blocked on information, it is avoiding a call.
Record the vote, including the dissent
Most committees at small funds vote informally, by nodding. The problem shows up eighteen months later when the deal has gone badly and nobody remembers who was uncomfortable and why. The partner who raised the exact risk that killed it gets no credit, and the committee learns nothing, because there is no record to learn from.
Recording votes solves this cheaply. In SamCIO each partner casts a yes, no, or abstain on the deal, with an optional written rationale attached. Each person holds one vote per deal and can change it, and every cast is written to the deal's time-stamped activity trail, so the sequence of positions is visible rather than just the final tally. That is the part worth having: a vote that flipped after a reference call is a genuinely interesting fact about how the committee reasons, and a system that shows only the end state hides it.
What this buys you is not governance theater. It is the ability to ask, two years later, a question that most funds cannot answer: when we were wrong, was anybody right, and did we listen to them.
Protect the abstain
An abstain is information and most committees treat it as an absence. A partner who abstains because the sector is outside their competence is telling you something useful about the quality of the coverage on that deal. A partner who abstains because they do not want to be the one who blocked it is telling you something about the committee's culture, and it is not good.
Make abstention a first-class option with a rationale attached, and read the rationales. If abstentions cluster on a sector, you have a coverage gap. If they cluster on deals a specific partner sponsors, you have a dynamics problem the votes are politely routing around.
Separate the thesis argument from the deal argument
A large share of unproductive committee time is two people arguing about a deal when they actually disagree about the fund's strategy. That argument cannot be resolved deal-by-deal, and attempting it means relitigating the same disagreement every time a similar company shows up.
When the discussion turns into "we should not be doing seed-stage hardware at all," stop and name it. That is a thesis conversation, it goes on a separate agenda, and the deal in front of you gets decided under the thesis as it currently stands. Committees that allow strategy debates to happen inside deal decisions end up with an inconsistent portfolio and a written thesis that describes a fund they are not running. A thesis-driven framework you can enforce is what makes this separation possible; without a written thesis, every deal is a thesis debate by default.
Small funds still need this
The common objection from three-person funds is that a formal committee is overhead when everyone is in the same room daily. The daily conversation is real, but it is not a decision record, and the thing you are missing is not the meeting, it is the artifact.
A two-partner fund can run this in twenty minutes a week: memo circulated the night before, recommendation stated, both partners vote with a sentence of rationale, outcome logged. That is not bureaucracy, it is the minimum required to be able to reconstruct your own reasoning later. Funds that skip it are not moving faster, they are just not writing anything down, and the bill arrives at the next raise when an LP asks how decisions get made and the honest answer is "we talk about it."
What good looks like after a year
The test is not whether meetings feel productive. It is whether you can pull up any deal from the past year and see the memo it was decided on, who voted which way, what the dissenters said, and what actually happened. That record is what turns a year of decisions into a fund that knows something. Without it, every partner who leaves takes the reasoning with them and the committee starts over.
For the structure of the document the committee decides on, IC memos covers the memo and vote record together.