Oxx publishes research on the backgrounds of the founders behind successful software and AI companies, noting that there has been a notable change in founder profiles between the SaaS era and the first wave of AI and machine learning businesses.
The investment firm analyzed 645 founders of 270 companies. Of these, 150 companies are in the software-as-a-service group for B2B companies that have achieved exits or stock market listings of more than $500 million, and 120 AI and machine learning companies are following a similar trajectory.
The findings suggest that academic and technical qualifications have become more prominent as the market moves from SaaS to AI and machine learning. The proportion of founders with a PhD increased from 6% in the SaaS wave to 18% in the AI and machine learning wave, and the proportion who went to a top university increased from 37% to 61%.
In contrast, MBA qualifications have become less common in later cohorts, dropping from 14% in the SaaS era to 4% in AI and machine learning.
This data also points to new changes in the current AI market. Among the companies that Oxx describes as the latest generation of founders building generative and agentic AI application-layer software companies for 2023 and beyond, the metric for Ph.D. completion rates has dropped from 18% in the early AI and machine learning group to 12%.
This suggests that the educational profile associated with early AI company builders may already be changing. New companies are focusing less on basic model development and more on software products built on top of existing AI systems.
Richard Anton, co-founder and general partner at Oxx and author of the white paper, said the change from previous SaaS to AI reflects the demand to build more technically complex products.
“It makes sense that PhD rates rose between the SaaS era and the beginning of the AI/ML era. In the SaaS era, you needed commercial skills to scale your company. In the early AI era, technical skills were the bottleneck for B2B success. These products and technologies are really hard to build,” Anton said.
founder mix
Along with formal education, Ochs found that having a technical co-founder remained one of the strongest common factors across both time periods. They found that 85% of successful SaaS companies have a technical co-founder, compared to 95% of AI and machine learning companies.
The continuity is striking because other founder characteristics also appear to have changed significantly over time. The report argues that although the specific technical skills needed in software businesses are evolving, founding teams still tend to require members with deep technical expertise.
Anton explained that this is one of the clearest patterns that runs through both generations of company founding.
“What unites both generations is the technical expert. 85% of SaaS teams and 95% of AI/ML had a technical co-founder. One thing remains the same: even if the nature of technical skills has changed, someone needs to be a technical expert when co-founding a team,” Anton said.
The report also links the increase in founders from top universities to the network and visibility those universities can provide. Oxx highlighted MIT, Stanford, Oxford, and Cambridge as examples of academic environments where AI and machine learning cohorts are more focused.
Anton said that attendance at these universities appears to be important even if the founders have not completed the course.
“Although not all of the founders had degrees, their attendance allowed them to build the right network, build relationships with co-founders, and attract investor attention,” Anton said.
regional shift
The study also shows that successful software company founding is not concentrated in the United States. In Oxx’s sample, the UK’s share of successful companies in SaaS in AI and machine learning increased from 1% to 12%, while Continental Europe’s share increased from 5% to 10%.
These numbers suggest a broader redistribution of software entrepreneurship towards European markets, particularly in AI-related businesses where university research power and technical talent are becoming more important.
Mr Anton said Britain’s academic infrastructure was a key part of that change.
“The UK is doing well because of world-class AI education and research at universities,” Anton said.
He added that Europe’s position has strengthened over several technology cycles as role models and funding for founders are no longer solely in the US.
“I’ve been investing in software for 30 years through various innovation cycles. I always saw that the UK and Europe had immense talent, but at the time the role models and capital were almost exclusively American. That is eroding and the UK and continental Europe have proven they have the talent, institutional and market complexity to build global winners,” Anton said.
investor signals
As AI tools become easier to use for investors, founder valuation patterns may need to change again, the report argues. If PhDs become less rare as technology becomes more accessible, investors may place more weight on founders’ judgment in identifying practical uses for AI and turning it into products.
Anton said this change will impact how venture capital firms read founder credentials.
“A big part of venture capital is pattern recognition, and this will continue to be important. But investors need to be aware that the patterns that are worth recognizing are changing, and they need to be aware of which signals to look for, especially when evaluating a founder’s talent,” Anton said.
