Recently, we announced the winners of the 2023 Datos Insights Insurance Technology Impact Awards. I’m writing a blog series examining industry trends as seen through the lens of the 2023 Impact Awards’ 65 case studies, which catalogue real tech projects that real insurers delivered to create real success. Today, I’ll be diving into trends from data and analytics projects across the industry.
Unlike digital projects, data tends to be a smaller category, because it so often captures projects that are intensive, multi-year efforts: things such as training algorithms to automate underwriting decision making or migrating an on-premise data warehouse to a cloud data lake environment. “Easier” data projects might be setting up a self-service data mart where business users can independently use reporting tools like Tableau to get their own analytics and insights.
Data projects are large, complex and difficult, but also crucial and unavoidable.
What’s clear from this year’s Impact Awards data case studies—and from the digital and core case studies, for that matter—is that practically anything an insurer would like to do ultimately comes down to effective data and analytics capabilities. Accurate, available data is the secret key to all the other capabilities insurers want to implement.
Take midsize property/casualty winner Mosaic Insurance (now a back-to-back Impact Award winner!). Mosaic implemented an underwriting portal to improve customer experience for high-end specialty lines customers. Creating the speed the team wanted required Mosaic to embed AI-based decision making capabilities within the portal so the system could quote, bind and issue policies automatically. What seems like a digital initiative on its face (“let’s sell specialty insurance online”) is actually a data initiative, because the algorithm is such a crucial component of the overall function.
To that point, the data used to train an algorithm must also be high quality, and any third-party data invoked in the new business process must be reliable. Fellow winner CNA is an example of the latter. CNA wanted to provide faster quotes, and its path to doing so was to build an AI-enabled automation solution to improve data extraction from forms.
These needs also extend beyond new business. Life winner Lincoln Financial wanted to improve customer experience, but what the team built was a holistic customer data view, because the core of that customer experience is being able to serve accurate information about accounts, on demand, to the portal or channel where the customer wants to view it.
See also: Achieving a 'Logical Data Fabric'
It’s data all the way down.
Whether an insurer wants to sell more, manage risk better, serve customers more effectively or differentiate itself from the competition with superior user experiences—all of it ultimately comes down to data and analytics capabilities.
Want to provide a superior distribution experience by providing instant quotes? Underwriting components are typically core, but you definitely need good data and good analytics.
Want to improve customer experience by pre-filling fields for your digital FNOL user flow? You need to be able to pull that information from a data lake.
Want to save on claims costs by more effectively flagging potential fraud or more accurately predicting claim severity? Pure data and analytics, which will have a clear impact on profitability.
Insurers, like everyone else, are rightly paying a lot of attention to generative AI and large language models like ChatGPT. But at baseline insurers need data that’s accurate, reliable and available, as well as algorithms that are trained on quality data and that produce decisions that can be trusted. More and more, data is core, data is digital, data is everything.
To check out all 13 data and analytics case studies, read Insurance Technology Impact Awards Case Study Compendium 2023: Data Initiatives. Interested to see what’s happening in the world of digital, core and IT practices? Find information about all of this year’s Insurance Impact Awards winners here.
