Use of Data in Product Development in Insurance: Present and Future
The amount of data that is available to an insurance company in various forms and variety is voluminous and increasing exponentially. Data comes in two forms- structured data which is available through the policy administration systems and unstructured data such as text inputs, audio files from call centre, claim adjuster notes, click streams, social media etc. Traditionally Insurance companies have been harvesting and analyzing only the structured data for routine business. There lies a huge scope in creating value through unstructured data if combined with structured data which can assist in deriving meaningful insights into the business models. Currently, data is captured at the time of policy purchase, claims interaction and other customer interactions within the policy administration system. This data is processed, analyzed and used as an input for product development. There are various other external sources such as industry reports published by the Regulator and other governing councils in Insurance industry which are used as a reference point. However, using the available structured external data source carries a risk of skewed behaviour due to difference of business models operating in the industry. For e.g. the data of a company primarily operating on the bancassurance model shall be very different than the company operating on other business models within same line of business.
For new product development in these times, there is a need to understand customer segmentation and their needs in detail through various external and unstructured data points and design a product accordingly. Below mentioned are a few means and ways which can be utilized for effective product development through the use of data :
• Claims and Fraud analytics- Text input analytics, social media analytics and predictive analytics can be used for detecting fraud claims and claims having litigation potential. The above data can further be refined and used effectively for customer segmentation and used for pricing. Presently better profiled customers also end up paying higher premiums due to frauds.
• Cross-sell and up-sell opportunities- By merging the direct customer profile data available through the policy administration systems and indirect customer data through social media, blogs and so on, call centres can increase the productivity by identifying the cross-sell and up-sell opportunities • Use of Telematics data- Today there are many mobile applications and tracking devices which can provide near real time information on the driving behaviour and pattern of a specific driver. There is a huge scope of analyzing this information and using the same for driver based underwriting T
• Credit History- It is an established fact that the credit history of the individual has a strong correlation with the insurance behaviour of the customer. This data can be merged along with the current insurance data thus strengthening customer segmentation.
• Wearable devices- there have been an uptake on the use of wearable devices for fitness tracking which can be used for good health incentives in health insurance.
• Social media for launch of product- Social media is a cost efficient and effective medium for introduction of new products in the market. It is lot easier to experiment with this approach with different customer segments. The above streams of unstructured/external data can be effectively combined with the available structured data. The insights derived can be helpful in new product development and for product customization towards specific customer segments. To aid product development, the above combined data also needs to be adjusted for various factors such as:
• Relevance/future relevance of the data
• Credibility of the data source
• Trend, transition, distribution and concentration of data
• Data abuse
The IT department of the insurer needs to play an important role in the whole exercise. There is a need for objective evaluation of investment in technology and infrastructure required for big data strategies. The future of effective and relevant product development lies in use of various streams of data through data collection and analysis and deriving insightful findings which can be effectively used for new product development.