Can Big Data technologies boost profits and enhance customer service in the auto insurance industry? Nick Turner, Enterprise Data Manager at U.K.-based insurance provider Markerstudy, joins us today to share his firm’s recent, award-winning successes with Cloudera and the Zoomdata visual analytics solution. Learn more in the upcoming webinar: Big Data Visual Analytics In Action: Increasing Revenue and Reducing Risk.
Markerstudy is a fast-growing provider of specialized auto insurance in the U.K. (where we call it motor insurance). We sell through thousands of agents and brokers, who request tens of millions of customized price quotes daily. The quotes must be customized because of widely diverse risk factors, demographics and coverage needs.
Every one of these pricing requests has to be thoroughly analyzed to produce a quote that’s competitive, appealing to our prospective client, and profitable for both our agents and us.
To do this efficiently, we built a web-based Insurer-Hosted Rating Engine (IHR)—a common portal for use by all the agencies and brokerages involved and by our own underwriting staff. Employing a cloud-based SQL Server database, the IHR gives Markerstudy staff control over the quoting process, allows easy integration with broker and data aggregator systems, and leverages real-time, third-party data to help create highly accurate prices and reduce fraud.
Success breeds data—in quantity. The IHR has been a great success—and has resulted in a tripling of the quote requests it must process daily. But constraints in computing capacity and data movement capability—along with the desire to mine the data for new insights—soon led us to explore the use of Big Data technologies.
Why Big Data? We knew that Big Data technologies could enable us to analyze more aspects of the data, for a range of purposes, in near-real time—and we wanted to leverage it in a way that our business users would embrace.
And so we launched our Big Data Insight initiative. After securing business sponsorship, our team began assembling Hadoop clusters using the Cloudera Enterprise and the Zoomdata fast visual analytics platform.
Rapid development. It took us only five months to have our solution in place, composed (in part) of a Cloudera Enterprise, Apache Impala (incubating), and Zoomdata for data visualization and analytics. Data sources and client systems include our IHR rating engine and its transactional, SQL Server database; enrichment services that provide customer credit and fraud histories plus vehicle, driver license and demographic information; and a range of search, analysis, operational and management tools. Impala is used as the primary query engine for Zoomdata, and for other ad hoc queries.
New analytics power—and new insights. We now analyze billions of quote transactions in seconds—instead of processing only a small subset over a seven-hour period. The time savings lets us respond almost instantly to quote requests, resulting in greater sales and enhanced service to both customers and agents.
We’ve also reduced claims costs by £5-million (USD$7.1-million) thanks to enhanced fraud detection. And because we better understand our customers and can provide instant, accurate pricing, customer cancellation rates have been cut in half.
Also, because analytic exploration and inquiry is dramatically easier through Zoomdata, we have identified previously hidden data patterns that help us improve business processes, delivering tangible, measurable results.
Markerstudy continues to grow rapidly, through successful marketing and frequent acquisitions. Now, with an enterprise-grade, Big Data infrastructure in place, we’re ready to support Markerstudy’s ongoing, rapid growth.
Learn more about what Markerstudy was able to achieve with Cloudera and Zoomdata in the upcoming webinar, Big Data Visual Analytics In Action: Increasing Revenue and Reducing Risk. The Webinar will take place on Wednesday, March 2nd, at 10:00 am PT. Reserve your spot today.
Nick Turner, Enterprise Data Architect, Markerstudy
Nick Turner has a career in data that spans more than twenty years, successfully delivering projects for organizations that include GlaxoSmithKline, DHL and EDF.He currently leads the Enterprise Data team at Markerstudy, overseeing the multi award-winning Big Data Insights project, responsible for the collection, analysis and visualization of hundreds of millions of records a day. He is a frequent speaker on the Big Data journey and lecturer in computing and data analysis.