Can AI Predict the Next Unicorn Founders?

The success of an early-stage startup largely hinges on the founders' ability to execute. Many pre-seed and seed-stage investors share this belief, placing a significant emphasis on evaluating the founders during their investment process.

However, after engaging in conversations with various investors, I've discovered that they remain sceptical about AI's ability to predict a founder's success due to the subjective and intricate nature of evaluating founders.

My curiosity to question these assumptions and determine if “AI identifies key traits of successful founders?” led me to create this article.

What I uncovered was AI can help identify successful founders due to the data patterns in personalities, and while it may not be perfect, AI can serve as a tool to investors or founding teams to increase their chances of success.

Remember when Neo was able to see the underlying code of reality through the use of a computer programme in The Matrix? Similarly, AI can help investors and founders "see the code" of successful startups, identifying patterns and insights that might be missed by human analysis alone.

This article delves into the critical role of high-quality founders in the early stages of a startup, explores the six personality traits that distinguish successful founders, and outlines how this knowledge can guide informed decision-making.

What makes a start-up succeed?

The leadership team plays a pivotal role in determining the success of any company, particularly in the pre-product-market fit stage. Even though initial products of startups might not always succeed, a strong and efficient team has the potential to adapt to new markets and develop new products. 

Notion serves as a shining example of a company that has undergone major pivots to achieve its current success. Their initial product, a visualisation tool for programmers, differs significantly from the Notion we know now. It took Notion several years of refining and pivoting before they unveiled the Notion that we use today in 2018.

In this process, Ivan Zhao, had to go through years of hardships and doubts to achieve product-market fit. He even moved to Japan to decrease his expenses! Ivan’s commitment and curiosity to understand user needs are what led to Notion’s success.

To further illustrate the importance of early teams, an analysis of 101 tech startups revealed that 23% of failures were attributed to not having the right team – the third most common cause of failure, surpassing running out of funds and lacking a product that satisfies market demand.

Since a company's success is largely determined by its talent, leveraging data and AI is crucial for investors and founding teams to enhance their chances of success.

A recent study by Oxford Internet Institute, University of Oxford, University of Technology Sydney (UTS) and the University of Melbourne used data and AI to uncover if we can predict a founders’ success, setting out to answer two key questions:

1. Which personality features characterise founders? 

2. Do their personalities, particularly the diversity of personality types in founder teams, play a role in startup success?

To find the answers, they delved into the data of 21,000 startups, utilising information from Crunchbase and social media platforms to formulate their hypotheses. They measured success as the occurrence of liquidation events through mergers and acquisitions or initial public offerings (IPOs).

Can AI identify the key traits of successful founders?

According to this study, the answer is yes.

The study revealed founders of successful startups possess personality traits that set them apart from the general population, and these traits play a more crucial role in determining startup success. They discovered a founder’s personality was more predictive of success than the industry (five times) and the age of the startup (two times).

In order to categorise personalities, they used the Big Five model. This favoured model includes 30 dimensions and 5 core traits: Openness to Experience, Conscientiousness, Extraversion, Agreeableness, and Emotional Stability.Studies have consistently demonstrated that a person's Big Five personality traits tend to remain the same throughout their life, even in the face of significant life-altering events such as divorce or job loss (Soldz, S. & Vaillant).

The study revealed six distinct personality types that characterise successful startup founders: Accomplishers, Leaders, Fighters, Operators, Experts/Engineers, and Developers. It turned out that these personality types had common characteristics in common, especially high adventurousness, intellect, achievement-striving, activity level and low emotional instability and modesty. 

Using personality data alone, their algorithm correctly predicted successful founders from employees with 82.5% accuracy.

Below you can find a diagram illustrating the common characteristics of the personalities.

The study further revealed that teams with diverse personality traits and complementary roles are more likely to achieve success. They found that the combination of personality traits among founders is important, and teams with a Leader/Operator, Expert/Engineer, or two Developers were more successful. 

One of the researchers, Prof. McCarthy states: “largely, founding a startup is a team sport and now we can see clearly that having complementary personalities in the foundation team has an outsized impact on the venture’s likelihood of success, which we’ve termed the Ensemble Theory of Success.”

These findings suggest that personality traits should be considered when forming a startup team. Reference the diagram below for the most successful founder team combinations.

Research limitations

While this research presents some interesting findings, it doesn’t come without its limitations, particularly related to the accuracy of data and methodology.

The data for this study is drawn from Crunchbase and social media profiles, which means that the study is only able to observe companies that have received venture capital funding or that are active on social media. This can create a selection bias by excluding companies succeeding without venture capital funding or lacking tech-savviness. Additionally, the data is not representative of all startups, as it is limited to companies that have publicly available information.

Additionally, the study uses language analysis to infer the personality traits of founders. This method is not without its limitations, as language can be used in a variety of ways and can be influenced by a number of factors. For example, founders may use more positive language when their companies are doing well and more negative language when their companies are struggling. Additionally, the study's approach to personality inference may not be able to capture the full range of personality traits that are relevant to startup success.

Despite its limitations, the study provides valuable insights into the role of personality in startup success that can be used by investors and founders alike. 

Practical applications of personality insights

This study serves as a valuable baseline for investors and entrepreneurs to further refine their understanding of personality traits and their impact on startup success. 

In depth personality profiling supports an informed decision-making process. Looking out for common personality traits, such as high adventurousness, intellect, achievement-striving, activity level and low emotional instability and modesty can serve as a valuable way to use these findings. 

To understand an individual's personality traits, you can either acquire the expertise of profiling, such as DISC profiling, or employ more accurate algorithmic methods. Individuals can create their own proprietary algorithms by collecting data on the personality profiles of founders and teams. This approach produces superior results, which is why more funds should be focused on collecting data. However, for individuals, this process is more resource-intensive and may not be the most efficient solution.

An example fund taking this initiative is Episode 1 Ventures, who have adopted a data driven approach, evaluating over 15,000 pitch decks and 5,000 founders to pinpoint what makes a successful startup.

Similarly, some of their key findings include:

• Multiple co-founders can significantly boost the likelihood of securing a Series-A round.

• Previous experience as a founder can double the probability of successful fundraising.

• Length of experience isn't a key factor; founders with less than five years of experience have equal chances of raising a Series-A or Series-B as their more seasoned counterparts, with 15-20 years experience.

Since adopting an AI-driven approach, Episode 1 has widened their top-of-funnel by 53% in just three months. 

The study's findings reveal the power of AI in understanding the personality traits that drive startup success. With AI-powered personality analysis, entrepreneurs, investors, and others can gain valuable insights to identify and support the teams destined to achieve great things.

If you want to dive into the study on a deeper level, you can find their full results here.

If you're enjoying this article, make sure to subscribe to my newsletter here where I post every Friday UK morning.