Overcoming the Insurance Onboarding Challenge with Knowledge Graphs
February 10, 2020 | By: Kathy O'Neil
With reports of approximately $4.5 billion in investments in 2018, insurtech’s impact on the insurance industry is steadily growing—especially for consumer expectations. Traditional insurers are struggling to accelerate their onboarding and processing times to keep up.
Customers now expect quotes, premiums, and policies to be processed faster than ever, which raises the bar for ongoing service delivery, too. The key to meeting these demands is leveraging public and non-traditional data sources. Companies that quickly integrate and access these sources can provide almost instantaneous policy creation. Semantic knowledge graphs combine such rapid data blending with fast data discovery to fulfill these needs across insurance domains. They enable organizations to meet consumer onboarding expectations at scale for consistently quick service delivery.
The Front End: Instant Onboarding
Traditional insurance onboarding processes are extremely lengthy. Customers are required to answer seemingly endless questions. In most cases the data entry process is manual; data are usually input into various silos. The entire process can take weeks, during which time potential customers are left waiting. In the meantime, the threat of their churning (especially in today’s insurtech age) is a risk many non-traditional providers can’t afford to take.
Modern boarding processes are nothing like these legacy ones. They typically involve consumers downloading apps on their mobile phones. Through these, they answer a limited number of questions—which usually pertains to information not readily found via public data sources. The result is near real-time responses for quotes, policies and premiums. Insuretechs like Lemonade and Hippo consistently provide this information well within 10 minutes’ time. The result is a much better consumer experience and shorter time to value for insurers, as well.
The Back End: Knowledge Graphs
The first way knowledge graphs support these front end benefits with swift processing on the back end is via data blending. These graphs standardize data of any variation or source in a couple key ways. They leverage data models that naturally evolve to include new sources. They also use uniform vocabularies and taxonomies so data is described in the same terms, regardless where they originate. Importantly, the data preparation work for these models and terminology is done upfront. Therefore, they can quickly ingest data and harmonize public data sources with internal ones for agile, rapid analytics resulting in policy creation.
Because of the uniform terminology describing data with business concepts, the data discovery process is much quicker with knowledge graphs. The semantic standards supporting this approach are machine readable, allowing business end users and automated methods to discover data. When underwriting an auto insurance policy, for example, organizations can access public sources for a prospect’s driving history. Once those data are integrated in knowledge graphs, machine learning techniques can use them to see which policy options pertain to them. Insurers can rapidly understand which consumer data relates to factors for determining policies. Moreover, these procedures can be repeated to automate transforming the selected data into target systems for analytics resulting in policy creation.
Ongoing Processing Benefits
Knowledge graphs improve the insurance onboarding process by drastically decreasing the time it takes, which greatly reduces the threat of churn. Their seamless data blending and fast data discovery support the insurance industry’s transition into the insurtech era. Furthermore, these tools are essential for the sustainable use—and reuse—of various data for the long term. Best of all, they offer these same advantages for ongoing claims processing, improving the customer experience even more.
You might also be interested in reading, “Perfecting Homeowner Insurance with the Internet of Things”.