At Metal, one of our focus areas has been in-depth research on the investing behaviours of venture investors. Across historical data, one prominent trend is that venture investors love to invest in patterns that they believe will deliver returns. For instance, investors that invest in HR Tech once or twice are significantly more likely to make additional investments in that particular area (relative to investors that haven't done any prior work in the space). This unfolds an important truth about venture investments:
When Investor X makes an investment in a specific vertical, that event serves as evidence that the investor views that space favourably. Put differently, investors that have made multiple investments in HR Tech are more likely than others to make further investments in that space.
Understanding Opportunity Spaces
In order to identify similar companies, one layer of complexity lies in understanding what exactly qualifies as "similar". As an example, a company that builds a hyperlocal marketplace for senior care is actually strikingly similar to one that develops a hyperlocal marketplace for gardeners.
Traditional sectors would categorize the former within healthcare and the latter within home care. In reality, however, both companies have a super similar business model (built around enabling users to earn money by delivering basic services within their neighbourhood).
In the long term, with the right quantum of training data, AI models will get really good at identifying similar companies. Alongside similarity in business models, another important layer of complexity involves the broader investing thesis that a given company falls under.
Collectively, similarity in business models and overlapping investing thesis come together to form what we refer to as an"opportunity space". Investors that have historically invested in the same opportunity space as a given company are ultimately the "most likely" financing partners.
Metal's Investor Ranking Models
At the very core of Metal's technology lies our investor ranking model. With the right quantum of training data, our model has the ability to identify the "most likely" investors for a given Company -- the model uses data points derived from on a unique combination of industry and user generated information. An early version of our ranking model is already live on the platform, enabling users to identify investors that are a particularly strong fit for their company and round.
Ultimately, we believe that the overall venture industry will benefit from AI models that can identify which investors are most likely to lean in on specific types of companies and/or rounds.
Founders that zoom in on the right type of "most likely" investors are often able to close rounds with a higher level of certainty than those adopting a "spray-and-pray" approach. All else being equal, pursuing high-relevance investors reduces the time it takes to close, improves overall odds of a successful close, and eliminates the need to talk to a very large number of investors.
