Ventiv Resource Library
Issue link: https://ventiv.uberflip.com/i/1454073
Data is traditionally scattered across the organization, in data silos and legacy systems. The ability to access and pull that data together in a coordinated, concerted effort can sometimes be a herculean task. Insurers rely primarily on structured data -- the kind that is highly organized, formatted, and searchable in relational databases. Unstructured data is just as the word says, there is no pre-defined data model which makes it hard to manage and use. And analysts estimate up to 80% of today's data is now unstructured. While its data that ultimately feeds the AI machine, too much data (or for that fact, too little data) can affect the anticipated outcome. While a lot of data is usually considered a good thing, a good portion of that data may be irrelevant or not useable. It is often hard to separate the important data from the "noise". On the flip side, too little data for an analytics model may not be enough to provide accurate results. BRAVE NEW WORLD | 3 Of course, data quality is a critical factor as well. As the old saying goes, "garbage in, garbage out". Many insurers lack standardized data definitions, metadata, and proper governance procedures when it comes to storing and maintaining data in their systems. While carriers suffer from these issues, there is still plenty of good data that can help drive the analytics initiative. And those AI initiatives are making a difference. Actionable Insights for Insurance If there was a silver lining to the COVID-19 epidemic, it was the fact that it changed old habits and created new ones. Most important of all, it became a driver for digital transformation. With mask mandates, social distancing and restrictions on large group gatherings, companies were forced to change work behaviors and practices. And this accelerated the need for digital adoption and adjustments. This digitalization disruption serves as a catalyst for the development of new tools and techniques around artificial intelligence, machine learning, data, and analytics. The main result: innovation.