Insurers are pouring resources into modernizing their data infrastructure, but Olivier Lafontaine, chief product officer at Equisoft, warned that many are chasing innovation without a clear business case. “Insurance companies are investing more and more into putting their data there, extracting data, cleaning their data,” he said. “Ultimately that results in better insights into what the market wants, into what the customer base wants, and then better products, better offerings.”
According to Lafontaine, the shift toward data lakehouse models - hybrids that blend the structure of traditional data warehouses with the flexibility of data lakes - is helping insurers unlock richer insights faster. These platforms are allowing companies to centralize structured and semi-structured data, powering more advanced analytics and faster iteration cycles on new products.
“It’s the integration of the data ecosystem,” he said. “The sharing of the data in a safe and secure way… makes it possible for the whole chain to be able to better serve the customer.”
Yet despite that momentum, Lafontaine cautioned that the insurance sector has a history of getting distracted by the latest technology wave without evaluating the return. “We’ve seen that before in the dot-com era,” he said. “There was a little bit less of that for the blockchain technology that we heard of a couple years ago. We almost don’t hear about this anymore.”
Lafontaine urged insurers to resist building from scratch or experimenting too aggressively with tools from big tech firms. “It’s really hard for insurance companies to go after startups, or try to use what Google produces, or Anthropic,” he said. “What seems to be working quite a bit is if you work with Salesforce, for example, or Equisoft… then you’re kind of getting the first mile done for you.”
Working with established vendors offers pre-built use cases and the benefit of lessons learned from others in the market. “Some companies are more advanced than others. They failed at certain things,” he said. “They bought our product, perhaps subsequent to that, and now they know that there’s certain things that work well.”
Lafontaine pointed to sentiment analysis in contact centers as one use case that has demonstrated clear value. “It doesn’t replace the call center person… but it really helps you getting on the call, knowing that the person is angry or is in a good mood,” he said. “You can even have recommendations for the agent… ‘This client is angry. Start with appeasing.’”
While enthusiasm around generative AI continues to dominate boardroom discussions, Lafontaine argued that real innovation hinges on rigorous assessment. “You have to be careful not following the trend for the sake of the trend,” he said. “It’s very easy to get caught by the excitement… but for everything that you’re going to get it to do automatically, it creates additional problems.”
He stressed that experimentation was only valuable if paired with strong measurement. “Measurement is the part that is so difficult to achieve with technology,” he said. “Everyone gets excited about the new technology.”
AI may accelerate data cleanup and improve analytics dashboards, but Lafontaine emphasized that insurers must stay focused on delivering business value. “We talk about AI; we don’t talk too much about analytics. That’s 10 years ago,” he said. “But because of AI, analytics is getting so much better… you can interrogate your data, you can do so much.”
