AI companies around the world raised more than $100 billion in venture capital in 2024, an increase of more than 80% compared to 2023, according to data from Crunchbase. This represents almost one-third of the total venture capital invested in 2024. This is a large amount. Money is pouring into many AI companies.
The AI industry has swelled tremendously over the past two years, filled with overlapping companies, startups that only do marketing but don't actually use AI yet, and legitimate AI startups that are diamonds in the rough. It has become. Investors have a hard time finding emerging companies with the potential to become category leaders. Where does it even begin?
TechCrunch recently surveyed 20 VCs who back startups building for the enterprise to find out what gives AI startups their moat and what sets them apart from their peers. More than half of respondents say it is the quality or scarcity of proprietary data that gives AI startups an advantage.
Paul Drews, managing partner at Salesforce Ventures, told TechCrunch that it's very difficult for AI startups to build a moat because the landscape is changing so rapidly. He added that he is looking for startups that combine differentiated data, innovation in technology research, and a compelling user experience.
Jason Mendel, a venture investor at Battery Ventures, agreed that the technology moat is shrinking. “I look for companies that have deep data and workflow moats,” Mendel told TechCrunch. “Access to unique, proprietary data enables companies to deliver products that are better than their competitors, and robust workflows and user experiences allow companies to deliver the engagement and intelligence that customers rely on every day.” can become the core system of
For companies building vertical solutions, having proprietary or hard-to-obtain data is becoming increasingly important. Scott Beechuk, partner at Norwest Venture Partners, says companies that can focus on proprietary data are the startups with the most long-term potential.
Andrew Ferguson, vice president at Databricks Ventures, says rich customer data and data that creates feedback loops for AI systems can make them more effective and help startups stand out. says.
Valeria Kogan, CEO of Fermata, a startup that uses computer vision to detect pests and diseases in crops, told TechCrunch that one of the reasons Fermata has been able to gain traction is because its model He said he thinks it's something that has been trained from both customer data and data. A product from our own research and development center. Kogan added that the fact that the company does all of its data labeling in-house also helps make a difference in terms of model accuracy.
Jonathan Lehr, co-founder and general partner at Work-Bench, added that it's not just about the data a company has, but how it can be cleaned up and put to use. “As a pure-play seed fund, we focus most of our energy on vertical AI opportunities that address business-specific workflows that require deep domain expertise, and that AI is primarily “It allows us to capture and clean data that was previously unavailable (or very expensive to obtain) in ways that would have taken hundreds or thousands of human hours,” Lehr said.
Beyond data, venture capitalists say they're looking for AI teams led by strong talent, teams that already have strong integrations with other technologies, and companies that have a deep understanding of customer workflows. .