Finding the next startup that will hit it big could be a calculation away.
TRAC, a San Francisco-based early-stage venture firm, developed an AI-powered model to predict the young startups likely to achieve a $1 billion valuation.
Dubbed by its cofounders as "Moneyball for venture capital," Insider's Ben Bergman got a peek at the model's methodology along with 30 early-stage startups it identified as potential future unicorns.
TRAC says startups flagged by the model have a 20% chance of actually reaching unicorn status. And while that success rate might not seem high, only a tiny percentage of young startups end up sniffing that type of valuation.
As Ben notes, TRAC's approach is a departure from how early-stage investors typically operate. Despite a proximity to high-end tech, VCs are usually less scientific when investing in young companies, relying on gut feeling, founder background, and personal relationships.
One of the most predictive factors for a startup's future success, according to TRAC's model, is who is backing it. That includes a group of 300 top angel investors and firms TRAC defines as "SuperForecasters."
(TRAC declined to disclose the SuperForecasters, beyond OpenAI CEO Sam Altman, but did highlight characteristics most of the angel investors in the group share.)
To be sure, finding a startup is only half the battle. Being able to invest in it remains "a monumental hurdle," TRAC cofounder Joseph Aaron told Ben.
Leveraging AI to help source deals is becoming increasingly popular in some circles. Swedish PE giant EQT started using AI-powered tools to make "cyborgs" out of its dealmakers.
Blending humans and machines is also popular in the world of trading. "Quantamental" investing strategies, or combining human-led fundamental analysis with quantitative models, have gained momentum over the years.
But finding the appropriate balance between the two sides, whether it be from a cultural or procedural perspective, is no easy task. And it might be a tough sell for VCs who pride themselves on sussing out young companies.
For example, what happens when the model indicates a startup that's a favorite of a fund's partner is actually a bad bet? Or vice versa?
No comments:
Post a Comment
Note: Only a member of this blog may post a comment.