Insurance and Big Data

Insurance companies work with customers to protect their assets. Daily they screen new applications, process claims and make necessary payments. Risk department of Insurance companies manage risk and fraudulent claims. Dealing with large amount of data about customer information, risk exposure, claims information, evidence details, reports is hard. Hadoop provides a centralized storage solution where all of the data can be stored and analyzed to serve customers in a better way and contain fraud.

Fast New Policy Screening

Insurance companies screen applications daily for auto, home, appliance insurance policy. They look at customer's credit history, income and other factors to determine eligibility.

When historical data is collected and stored in Hadoop, Underwriters can do proper data analysis and evaluate risk for desired coverage.

They can validate application details for fraud rings and automate the process to reduce costs.

Underwriters can build price risk models in Hadoop and evaluate new customer's risk to appropriately price the policy and coverage.

Insurance Agent Performance

Agents work with customers for their claims and policy related questions. Their actions are captured as interaction events. These agent events provide great insights about agents, their performance, effectiveness etc. Customer care team can analyze the collected interaction data to gain following insights:

  • How many requests agent/agents handling
  • Average call/chat duration, distribution by day, weekday, hour etc.
  • Type of requests - address change, password reset, account closure.
  • Geo distribution of agents, customers
  • Effectiveness/Performance scores
  • Average customer satisfaction

Fraud detection

Insurance companies can store following data in Hadoop to co-relate and detect fraud:

  • Customer data - age, location, income etc
  • Claims - current claim and prior history
  • Demographic information
  • Police & Adjustor reports
  • Social media including news
They can build machine learning models to detect anomalies and graph based models to detect connected crime ring.

Proactive Risk Predication

Insurance companies can collect lot of external data feeds like weather, sensor, social media, news in Hadoop.

They can build models to detect or predict risky conditions or events which could pose danger to customers and inflict damage to them or their property. These models can alert operations teams and customers to contain risk.

Customer 360

Insurance companies depend on their customers for business and their success depends on them. It is very important to understand customers, their needs, satisfaction to build long lasting relationship. This drives more business and creates brand value. To understand customers it is essential to know:

  • Who are the customers, where they come from, what they do
  • How customers interact with the company
  • What is their journey with the company
Big data solutions enable companies to collect customers data from all touch points (Claims, payments, police reports, adjustor reports, social media etc) to build 360 view and allow them to know and serve their customers well.


Knowing customers better will drive right targeted campaigns, which will result in better customer experience and conversion. Big data based models will allow creation of customer data sets to know who are the customers, where are they location, what they are looking for etc. It will also lead to identification of cross and up-sell opportunities.