The survey was administered to 391 VC investors across the US, Europe and Asia during the winter of 2018. According to the findings, respondents across all investor types and geographies rely on a mix of data and personal networks to source and evaluate investments. While the majority (86%) of respondents believe data is important when evaluating investment opportunities, late stage VC investors based in the US and Asia were the most likely respondents to leverage data to inform all investment decisions. Despite growing interest within the asset class to finance the machine learning (ML) and artificial intelligence (AI) category—$24.6 billion invested in 2018—only 8% of investors believe investment decisions will be fully automated in the future. More than 85% of respondents believe there will always be some element of intuition involved in VC dealmaking.
“There’s no question that the availability of data has led to whole new way of doing business. The accumulation of data over the years combined with faster computing power has allowed AI-powered innovations to transform entire industries,” said Steve Bendt, VP of Marketing at PitchBook. “As VCs flood capital into the ML/AI sector, we wanted to understand how VCs themselves leverage data and machine learning techniques in their own investment sourcing and decision-making process. Our survey shows strong adoption of data to inform investment decision-making and a growing appetite to increase usage. While the majority of respondents believe VC investing will always involve the human element, there’s enthusiasm to explore how machine learning can automate traditional VC.”
Personal networks and data are top resources for deal sourcing
The top three most valuable resources for sourcing and evaluating investment opportunities are personal networks (82%), inbound leads (44%) and financial databases (36%). Angel and early stage VCs are roughly 20% more likely to lean on personal networks for investment decision-making than late stage VC and CVC, which show the strongest demand for financial databases, 20% and 22%, respectively. Geographically, personal networks remain the top resource for sourcing and evaluating deals. Asian VCs are the most likely group to leverage data for this purpose, with 20% of Asia-based survey respondents citing financial databases as the most valuable resource, compared to 15% in the US and 10% in Europe. In practice, the most common use cases for leveraging data are financial modeling (20%), refining investment thesis (19%) and sourcing investments (19%). Among the top reasons for missing out on a promising investment opportunity are disagreement on terms (32%) followed by insufficient data (17%) and slow decision-making (15%).
Investment decision-making will always balance data and intuition
More than one-third (38%) of all respondents use data to source all venture capital investments, whereas 48% claim data informs some investment decisions and 9% don’t leverage data at all. Late stage VCs show the strongest appetite for data-driven investing, with 50% citing data as extremely important and the primary resource for evaluating and sourcing all investments. Similarly, 46% of Asian firms cite data as extremely important, compared to 37% of US firms and 33% in Europe. Of all investor types, angel investors are least likely to leverage data – 14% do not use data in the decision-making process. This isn’t surprising given the limited availability of data on niche and nascent industries, which is a common investment preference for this investor type. In addition to data, the majority of respondents (69%) across all geographies and investor types agree that both data and intuition are important when evaluating VC investments. A higher percentage of CVCs (23%) state data alone is the most important factor, while angel and early stage VCs are more likely to state intuition, 11% and 14% respectively.
VCs are bullish on machine learning, but are proceeding with caution
Nearly two-thirds (64%) of all respondents do not currently leverage ML technology to inform venture investment decisions; however, of this group, 50% plan on adopting and increasing usage in the future. Taking a closer look at investor types, more than one-quarter of CVC investors have already adopted ML, the highest adoption rate of all investor types. While this group is also eager to increase adoption (47%), early stage VC investors were the most enthusiastic about expanding this capability (55%). Early stage investors were also the most optimistic about relying on ML to fully automate investment decisions in the future, with 11% agreeing with this trend. Asian investors had the highest percentage of ML adoption, with 29% claiming to currently use the technology in decision-making. Although, 24% of US firms also leverage the technology, investors in this region displayed the most interest in increasing usage (51%). Regardless of current usage of ML, 85% of respondents believe VC investment decisions will always involve some element of intuition. Only 8% of respondents feel strongly investments could be fully automated in the future.