Data Analytics in Insurance

I attended the IBM World of Watson event last October. It was a great occasion to learn about the latest developments in data analytics and cognitive computing, to discover the IBM solutions for Financial Services and to meet with representatives from the Banking & Insurance industries.

Before highlighting how data analytics can be leveraged in Insurance, I would like to share 3 general observations:

[1] There is no single definition of Cognitive Computing. People agree broadly on the concept, but have different perspectives depending on their background. My definition:

cognitive computing is about leveraging the power of data analytics to create natural human-machine interactions, to extract knowledge from data, and ultimately to improve business decisions.

I divide Cognitive Computing into 3 categories:

  • Information capture (vision recognition, speech to text, …)
  • Information processing (NLP, clustering)
  • Machine learning (deep learning, predictive analytics)

[2] “Algorithms” cover a wide range of techniques from simple statistical reasoning, to machine learning and artificial intelligence. So when people try to sell you their “big data”, “data analytics” or “artificial intelligence” solution, it is worth digging to understand what is behind the words.

[3] Augmentation was a key theme: “intelligent” robots won’t replace humans completely, they will AUGMENT humans. It is not just a buzz word but something real. For instance in 2016 the songwriter-producer Alex Da Kid composed his song “Easy” with the help of IBM Watson. Watson analyzed both the sounds and lyrics of all the billboard hits for the last decades and was able to identify patterns. Alex used this output to refine the analysis and select some of the song’s themes. The song was the result of interactions between the human and the machine.

Now back to Insurance. Data analytics can be leveraged at different steps of the Insurance value chain:


  • The analysis of customer segmentation and behaviors (predicting the churn to increase retention, anticipation of life events, etc.)

Sales & Services

  • Personalized advice to customers (product recommendation, cross-selling)
  • Support of contact center agents

Risk scoring & underwriting

  • Dynamic modeling of risks (combination of connected sensors with advanced analytics)
  • Fraud pattern detection

Operations & Claim management

  • Cognitive Automation: cognitive computing coupled with robotics process automation

For more details on IBM products and some of the use-cases, please check the IBM Solutions for Insurance.

I learned a lot during a short period of time. It was a great experience. Thank you Susan for the opportunity and for the perfect organization!

Tagged with:

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.