Team or tech?
One of the standing jokes of venture investing can be illustrated by describing your investment focus evolution in three stages:
In the first, initial stage, the beginning years of investing, all startup investors start out investing with a clear focus on the founder team and their qualities. The main reason is that everybody you trust says it is the most important factor to address.
In the second stage, and the following years, as you immerse yourself in the startup world, attend tech and startup events and conferences, you nerd out more and more on the trends of tech, new business models, growth metrics, market analysis, and all the “cool” nerdy startup lingo.
Then for the third stage, as exits are getting closer, the true performance of your investments is revealed and your focus is switched back on assessing teams… and you tell everyone you meet. Startup investing is about investing in people and great founding teams. The circle is complete.
This portrait of the investor journey is also funny because it’s true.
The VC competency paradox?
It's hard to find a VC who does not agree that the quality of founding teams is crucial to get right. Paradoxically, and at the same time, founding team analysis is relatively little understood. And what constitutes a great team is less than agreed upon.
The paradox is even clearer when you engage in a dialogue with any experienced VC. The conversation can last for hours when it comes to tech trends, new market opportunities or fund modeling, but is equally short when it comes to founder team analysis. Sometimes it boils down to a short “I liked the energy and the founder's background”, or “I felt they had the right attitude”.
Since the fifties there has been a magnitude of psychological research on team performance, group dynamics and behavior, personality traits, and team functioning. The popularisation of this knowledge when it comes to applying it throughout the recruitment industry, or for team and organizational performance is yet to be represented in startup screenings and DDs. At least not fully.
Characteristics of successful founders: What is revealed in the data?
One of the more compelling analyses of the characteristics of winning founders - is compiled in Ali Tamaseb's book “Super Founders - What data reveals about million-dollar startups”. By crossing data from 200 unicorns and their 500 founders, with the control group of some thousand “regular” startups, he identified some key characteristics. Some of these are mythbusting, and some are confirmations of popular belief. And some, like this list, just confirm variables that have low correlation:
- Age: For instance if you believe age matters, the age of successful founders is evenly distributed from 25 to 60, with a few more in the middle.
- Education and elite universities: None of the successful companies was founded by first-timers who just finished a university degree. A longer educational background is a bit better than a shorter one, but not much, and being part of an elite university alumni does not correlate.
- PhD: What correlates though is founders with PhDs tend to succeed more often, and are way more regular than the famous school dropouts.
- Tech CEO: Successful tech startups just as often as not have a CEO with commercial or other background. Although tech knowledge is mandatory.
- Industry knowledge: Neither long-term employment nor background from the startup industrial vertical is highly correlated with success.
One common trait - the entrepreneurial drive
The not-so-surprising, and still valuable insight that can be read from the data is one core component and a common trait. Successful founders and founding teams simply have a history of taking the initiative to build things, and getting people to like, join and use it. Not necessarily in the shape of previous startups. Some have formed new communities during their studies, built NGOs, driven initiatives among friends or larger groups, they fundraised for a cause, or they simply have tried to build companies before, and they seldom succeed on the first trial.
To conclude, if one is to look for a valid signal, it should be their history of initiatives to build.
The successful founder personality - does it exist?
Talking to VCs there are a lot of theories out there, but none are yet supported by data. Personality traits do not correlate directly, and some of these myths could be killed:
The Troubled Founder Bias
By definition, out of the extraordinary achievements is accomplished by people with extraordinary motivation, which often comes from life experiences. If you have the drive it takes to be a founder, it could be reasonable that the hunger needed is coming from scars and or holes that need to be filled by success. Looking at the most famous founders this certainly often looks like the case. At the same time, the opposite is also true. To be successful, you need confidence in knowing a multitude of hard and soft skills often derived from the ability to explore the world from the place of a harmonic upbringing.
Troubled founders might more often end up in some kind of newsworthy drama that attracts more attention, and therefore is easier noticed. There are several such attention biases which affect everyday theories, and are not supported in the data.
To be able to learn fast, be data-driven, understand the tech trends, listen to the market, create customer trust, and build a strong team, are not defined by identified personality traits.
At the same time, there is no way around having general intelligence and ability to solve multidimensional problems which go a long way.
The unstructured data of humans - a perfect match for AI?
Humans are not always very good at understanding humans. Biases and egos come in the way. Everything from personality to upbringing and past experiences, to societal culture and situational elements is always at play.Too much unstructured data, multiplied by physical and social variables, makes it hard to analyze. In short, a perfect match for AI.
To augment the understanding, assessment, and analysis of founding teams will - by human nature - never be perfect, but it can easily outcompute human analyses without AI and data.
For us, the first step applied when screening companies is to look for the well-known proxies and indicators commonly used by VCs. Such as previous founder and entrepreneurial experience, relevant educational background, leadership experience, and general career path. Based on this you can also identify an eventual entrepreneurial history.
Step two is to include a more in-depth founder team analysis that requires input from the founders and validate it based on in-depth psychological analysis of founder teams. Right now we are seeing some promising approaches to integrate into the Northstar platform as we build more DD functions into the screening process. This includes mapping out not just the individual founder traits but the complete team picture including soft skills ranging from stress resistance, resilience, growth vs fixed mindset, self reflection, in addition to grit and passion resistance per team member.
Step three is to profile teams by going from action data to founder team characteristics to predictability models. Right now, that requires models yet to be developed…
Through augmented team analysis, experienced and analytical investors - already meeting several hundred founders every year - really get the augmented superpowers needed to pick the winning founder teams.
If you'd like to stay up-to-date on the latest advancements of AI for Impact Venture Capital, sign up for the newsletter here.