The investment decision is the product of a VC. Hedge funds productize decisions through data and algorithms, and banks do it with rules and scorecards. However, most VCs rely on a clunky process and partner meetings to make their decisions. The argument is that VC is an artisanal business and that data is sporadic and unreliable. This mindset makes the VC “decision product” vulnerable to bias, inconsistency, and noise - three deadly features in decision making.
On the other hand, founders are the consumers of the VC decision product. Most VC induce founders to provide information about their project, incrementally, starting with a slide deck. Then the VC and founder go through follow up meetings, questions, and data requests. The learning process for the VC is unstructured, so decisions can be inconsistent. Throughout the process, the founder has limited visibility on the VC’s decision, which can be frustrating.
Improving The Product
The problem starts with the pitch deck. Decks optimize for visual appeal vs. content depth, which isn’t an effective way of explaining a startup idea. Founders spend time reducing complex concepts into a few slides. This can make a pitch memorable, but short of details required for critical analysis by investors. A better approach would be for founders to explain their vision in long-form text, and for investors to prescribe information for founders to provide (if available). This way investors can establish a point of view for the qualitative and quantitative aspects of a startup that comes from analysis, and not just flash intuition.
This is where biases start to kick in. The founder’s looks, communication style, and affiliations can influence the VC to “like”, or “dislike” them. It’s important to work with people we like, but liking someone doesn’t make their startup a good investment. Consciously, or not, the VC may label the investment opportunity within a range of “good”, or “bad”, and determine the next steps based on it. One way to potentially reduce bias is to have a standard ranking method. For example, a VC could rank each of the founders during the meeting, across predefined dimensions.
Analysis & Investment Memos
A VC “liking” a startup typically leads to Q&A with founders and research. Most of this data exchange isn’t monitored and captured in notes, or emails, if at all. The output is a memo that partners consume in order to decide on the investment opportunity. This approach is a petri dish for confirmation bias, which means ignoring information that invalidates the beliefs that made the startup likable. One way to potentially reduce bias is to record all raw data and letting people make their own analyses. Another is structuring the data discovery and aggregation process in such a way that forces the good, bad, and ugly to be exposed in the memo.
Groupthink is the epitome of poor decision making. Partner meetings are when founders pitch VCs, deals are discussed, and decisions are made. The issue with deciding as a group is that poor decisions can be reinforced by the group instead of challenged. Focusing on making efficient decisions by consensus, or one person dominating the decision process are red flags. One way to mitigate this is to have people independently document their assessments before group discussion. This helps uncover key areas to focus on while limiting the group’s influence. Another way is to apply a technique called red teaming.
VCs are often called “pattern matchers.” In general, humans save time by making decisions based on patterns. Founders have been trained to explain startups as the “Uber for x” to make the matching easy. Similarly, some VCs look for “comparables” when analyzing investment opportunities. However, history shows outlier investment outcomes rarely fit existing molds. And if venture is a slugging game, then “pattern matching” may be a disservice to the practitioner. Conversely, VCs can benefit from seeing things for what they are, independent of what they look like. Reasoning from first principles has been talked about enough, so I don’t need to elaborate. However, applying this kind of thinking is hard as it demands lots of energy and self-awareness.
Investor’s fear of missing out (FOMO) typically makes companies expensive. I think about this in terms of skill and luck. Paying more for something without lowering the expected return requires a) more skill, b) more luck, or c) both. The same thing goes for reducing the amount of information and time for making investment decisions, but perhaps the relationship is logarithmic vs. linear. In VC, investing because of FOMO is the same as outsourcing your core product. And outsourcing your core product is the same as having no product. It doesn’t mean that a hot deal is a bad deal, but there’s a difference between FOMO and bearing higher risk based on thorough analysis. The point isn’t to ignore emotions when deciding, but to observe them carefully as they may provide useful information. A decision journal where emotions are written down and explored with others can help gain insight into one’s psychology and blind spots.
Typically, a VC’s decision process is a black box to founders. Why this is the case is beyond my understanding. This may just be the consequence of a manual, unstructured process, or that providing details demands time and potentially eliminates optionality for a VC. For example, giving founders feedback that’s hard for them to hear may cut out a VC from participating in future rounds. Anyhow, founders invest significant time educating VCs and often don’t get any value in exchange. A better approach would be a) making the decision process transparent to founders and providing updates at every step, b) sharing the analysis work product even if the decision isn’t to invest, and c) providing specific reasons for the decision in the form of useful feedback. Of course, people can be wrong, so qualifying any feedback given, or received is critical.
Decision Traceability, Learning & Calibration
Analyzing data and decision patterns can lead to insight. In VC it’s easy to be dismissive about this given the long feedback cycles. The average time lapse from first investment to exit trends around 11 years for early-stage VC. But if VC is a lifetime practice, then tracking decisions and looking back to calibrate still applies. The other argument is that luck plays a key part in VC, so a bad decision can lead to a good outcome, and vice-versa. But if there’s no measurement, how will a VC ever be able to understand if their product is good, let alone make it better?
This post isn’t a formula for success in VC. All of this stuff is easy to understand, but hard to practice. My point is that there’s a way to improve the “VC product”, and that starts with how VC’s decide.
Thanks to Arjun Sethi for reviewing draft versions of this post.
A good fundraising founder experience typically goes like this:
🦸 Founder: “I need to raise my seed round. I’ll put a deck together and ask friends for introductions to top investors.”
[friend makes intros to VCs]
🕵️♂️ VC: “Friend (BCC), thanks for the intro to Founder. Founder, it’s great to be connected. Would love to learn more about what you’re building. Is there a pitch deck you can share in advance of a call?”
🦸Founder: “VC, nice to connect. Sure, see deck here. Can we chat next week?”
[meeting takes place, and the question iterations begin]
🕵️♂️ VC: “Founder, it was great meeting and learning about your company. I have the following questions…”
🦸Founder: “VC, good questions. See my answers below. Happy to chat more.”
[founder wonders when VC will make an investment decision]
🦸Founder: “VC, I wanted to check in and see if you have more questions and next steps?”
🕵️♂️ VC: “Founder, sorry, it’s been a hectic couple weeks. Can you meet with my partners?”
[2nd meeting takes place, the questions continue, and the founder still wonders about the decision, and so on… You get the point.]