Technologists, business executives, and investors seem to be in agreement that Artificial Intelligence and Machine Learning represent the future of software. While these technologies are showing great promise and some very large AI-based companies are likely to emerge, there are important structural differences between traditional SaaS and AI enterprises that have implications for the latter's scalability and revenue model.
As a16z notes, AI companies tend to have lower gross margins due to heavy cloud infrastructure costs, humans in the loop, scaling challenges due to edge cases, and weaker competitive moats due to commoditization of AI models and data.
We have heard these challenges arise time and time again in the insurtech space from both founders and their customers. Every insurance company processes new business submissions, underwrites policies, and manages claims slightly differently. Therefore, each customer implementation requires new workflows and data sets, which creates evolving, time-consuming edge cases to solve for.
Certain customers insist on retaining data and IP for their specific implementations, which unfortunately (for AI companies) means some enhancements are not made available across clients.
Both investors and customers want to pick the AI startups that will endure. So what should they be looking for in startups that serve the insurance sector?
Key characteristics include (i) user-friendly applications that are designed expressly for the insurance industry and that solve industry-wide problems, (ii) demonstrable ROI and (iii) actual customers. If an AI startup is able to overcome the initial challenge of convincing a large (typically deliberate and risk averse) insurer to invest time and share proprietary data, the odds are that insurer will remain a customer for years to come.
We have noticed in many cases that AI companies simply don’t have the same economic construction as software businesses. At times, they can even look more like traditional services companies.