The raft of pressures implied on incumbent insurers – customers’ changing expectations, disruption of the value chain, emerging technologies – calls for the capability to turn data into a strategic asset. That’s exactly what this Nordic insurance organization strived to accomplish as it consolidated its information in a data lake, setting the stage right for utilizing AI and analytics for business purpose.
Insurers that don’t focus on data-driven decision making will soon be overtaken by rivals that do. A data lake could thus be one of the most essential steps for an incumbent firm to start using data strategically. Why not traditional database management? Because it relies on using rigid technology infrastructure to capture mostly internal data in predefined formats, and this has become insufficient.
What is needed in the fast-moving industry and its eco-system, is the possibility to capture diverse types of data, encourage experimentation with data, generate comprehensive insights, and integrate data with decision-making. Mostly, as to this specific Cognizant client, the key drivers behind a data lake are all business-oriented. The company’s overall ambition is to become a customer-driven digital insurer, with the ability to predict and address the customer need before they express it, regardless of channel.
Becoming a digital insurer
The company wanted to drive and develop new business opportunities through data and machine learning, transfer to a more continuous cycle of “usage-based insurance”, explore and learn from the vast data sets in claims or customer behavior and improve customer experience through real-time processes centered in data and analytics. It also wanted to move away from a siloed country level perspective into a more Nordic and proactive one.
A data lake foundation held the potential to provide all that. In general, data lakes are low-cost data storage environments that use commodity hardware and an integrated technology stack of open-source data and analytical tools. To the company, a data lake meant a next generation data engine that could capture any type of data, generate insights via advanced algorithms and integrate these insights in real time with the core processes. It also provided a platform for experimenting with data to build new products.
Detecting customer churn
To begin the journey and to prove the business value, the insurance company decided to execute a minimal viable product (MVP) and to build a churn prediction model for their large portfolio of commercial customers. The success criteria for the MVP was carefully selected in the areas of ability to use diverse data sources, democratization of data experimentation, prediction accuracy with machine learning, integrating the system-generated predictions with systems of interaction and organization design and change needs to adopt this in business.
The MVP was successful, and the company is now looking to expand the program and include more areas, such as product & pricing, sales optimization and improvement of the customer experience while interacting with the company.
Supporting from A to Z
Why was Cognizant chosen as a partner then? Because of the vast skill set and the capability to handle the whole value chains – from strategy, digital use cases, implementation and maintenance to support. Technology wise, the data lake and machine learning space is complex and fast-evolving; a project of this size normally requires integration of several open source and commercially available technologies and applications, as well as integration with legacy systems.
Yet, a data lake is not to be viewed as an infrastructure issue, but rather a business enabler. And a data lake project can be quick; with an end-to-end MVP results showing business impact within approximately three months.
Any advice to those wanting to embark on a similar journey? Cognizant’s experience from several other data lake projects, shows some distinct make-or-break factors: Get the internal organization on board and put efforts on democratizing the use of data with modern techniques. Be ambitious; pick business drivers that will make an impact as opposed to simple low-hanging fruits which sometimes demand similar effort but don’t bring sharp results that can enthuse the business. Also find the center of gravity and pinpoint the accountability of building a platform which in turn enables business teams to experiment and build data products on their own – otherwise it won’t happen.
To learn more about Cognizant’s insurance solutions, please visit our insurance page.