Investment committee

How an investment committee works

On a Tuesday morning a partner presents a Series A deal to four colleagues. They read the memo the night before. Forty minutes later the deal is approved, approved with a smaller check, sent back for more diligence, or declined. That meeting is the investment committee, often called the IC.

1. Pre-read. The deal team circulates the investment memo and the due diligence findings a day or more before the meeting.
2. Presentation. The sponsoring partner summarizes the case, usually in ten minutes or less, because members have read it.
3. Questions. Members test the key assumptions and the risks. This is where most of the time should go. A common one is "who do we know who has worked with this founder?", which relationship intelligence answers before the meeting rather than after it.
4. Vote. Under the fund's rule: unanimity, majority, or consensus with vetoes.
5. Conditions. Approval often comes with conditions, such as a confirmed reference, a lower valuation, or a specific board seat.
6. Record. The decision, the vote, any dissent, and the conditions are written down and kept with the memo.

Committees often see a deal more than once. Many ICs act as an early filter as well as the final vote. What they weigh differs by asset class. In a 2020 survey of more than 200 private equity managers, Gompers, Kaplan and Mukharlyamov found the business model ranked as the most important factor in new investments, followed closely by the management team.

How VC investment committees vote

The survey of 885 VCs by Gompers, Gornall, Kaplan and Strebulaev asked funds how they make initial investment decisions. The answers vary more than most founders expect.

Smaller funds, healthcare funds, and funds outside California were more likely to require unanimity. The right column is a reading of the incentives each rule creates, not a finding from the survey.

What investment committees look for in a memo

Committee members are checking whether the deal team has done the work and whether the case survives pressure. In practice they look for:

• A clear recommendation and terms in the first paragraph.
• The few assumptions the return depends on, each marked as verified or not.
• Evidence on the team, because VCs in the same survey ranked the management team as the most important factor in selecting deals.
• Risks with diligence findings attached, not a list of generic hazards.
• How this deal compares to others the fund recently approved or passed on.

The last point is where many committees are weakest. Without a shortlist or ranked view of the current pipeline, each deal is judged alone, and the bar moves with the mood of the room.

Worked example: separating scores before discussion

Alder Bay Capital is a fictional four-partner fund with a unanimous voting rule. It noticed that the sponsoring partner's enthusiasm set the tone of every meeting. So it changed the process: each partner now scores the deal privately from 1 to 5 on team, market, and terms before the meeting starts.

For a fictional fintech called Ledgerly, the private team scores come back as 5, 4, 4, and 2. The spread prompts the first question: what did the partner who gave a 2 see? She had called a former colleague of the CEO, outside the reference list, and heard about a failed product launch the memo did not mention. The committee sends the deal back for two more references. It is approved a week later with the risk written into the memo.

Collecting scores before discussion is a simple form of calibration. It shows when partners disagree, which a spoken round of opinions tends to hide. Over a year, the fund can measure how often partners agree, the idea behind inter-rater reliability.

Common investment committee mistakes

• Seniority sets the order. If the managing partner speaks first, others adjust. Collect views in writing first or have the most junior member speak first.
• Presentation eats the meeting. If members have not read the memo, the meeting becomes a pitch and the questions never come.
• Dissent goes unrecorded. A no vote that is not written down cannot be checked later.
• Rubber stamping. If the IC has approved every deal it saw this year, the decision is being made somewhere else.
• Handing the decision to a model. Scores can inform the vote. The partners own it. That is what human in the loop means in practice.

How ScoringFactory approaches the investment committee

ScoringFactory gives the committee a score for each founder and company against the fund's own bar, learned from its past yes and no decisions, with every point cited to the record. Members can see where a deal ranks against recent ones and check the receipts behind it. The partners vote. See the venture and private equity use cases.