Introduction

Data mining, also known as knowledge discovery, is a rapidly developing trend in knowledge management. Once been used by scientists, data mining has become a tool for manager, analysts and experts in commercial and government structures all over the world.

Data mining techniques operate with large volumes of data. The accumulated information is placed into databases, also called data warehouses.

Depending on the field of data mining application, the technique is defined as a data warehouse processing or as a set of processes, related to knowledge discovery.

Definition

Â

• Data Mining (DM) is a set of processes related to analyzing and discovering useful, actionable knowledge buried deep beneath large volumes of data stores or data sets. This knowledge discovery involves finding patterns or behaviours within the data that lead to some profitable business action or to new discoveries in science. Obtained key variables can be used to build predictive models for decision making.
• Data Mining (also known as knowledge discovery) is a process of automatically searching data warehouses for hidden patterns and behaviours. Main data mining techniques are: statistics, machine learning, pattern recognition.
• Data Mining is an analytical technique, based on various disciplines machine learning, pattern recognition and visualization, modelling, evolutionary computation, statistical analysis etc.
• Data Mining can be  defined as a process of information extraction. It’s aim is to discover hidden facts contained in databases.

Â

In general terms ‘Data Mining’ is a process of “extracting hidden information” from the available databases and presenting that information in an organized and systematic manner. It will involve the use of sophisticated statistical and mathematical techniques such as cluster analyses, automatic interaction detection, predictive modelling, etc. to predict customer behaviour and suggest information trends of customers.

It is to be seen that the results of data mining are different from other data driven business processes. While in the case of customer data interactions, nearly all the results presented to the user are things that they knew existed in the database already. Data mining, is an operational strategy, to promote market & sale of the company, one the Companies try to capture information from a customer every time he comes in contact with any of its departments.

Information access not only helps an organization realize new opportunities and potentials, but also provides far reaching benefits from knowledge management. Organized collection of data generates a database, where each record is a set of data elements and facts. Organizing the data will lead to generation of information, which is required in applying to customer’s centric decision making. “One manager may access this data base for knowledge discovery while the other may access this for knowledge deployment.”

Data Mining in Insurance is emerging as a powerful area in insurance selling process, no company can deny it’s immense significance. Research has shown those insurance companies who have neglected the significance of data mining in insurance, they are nowhere in the insurance business. It is a fact, data can play indispensable role in selling insurance by devising proper marketing strategy and plans.

The fact remains same that the importance of data for the organization of selling insurance, must understood. Actually data refers to a series of facts or statements that may have been collected, stored or processed but not organized or placed into context.

The sheer magnitude of data available in large organizations serves as a strong base for companies desiring to be ahead of competition as well as for emerging organizations in departments. The prime touch points being the

Â

1. customer purchase call;
2. on-line enquiry;
3. service related call;
4. complaints generated by existing customers etc.

Â

The information generation will include the personal characteristics of the customer, their preferences and choices of products, usage patterns and purchase histories including a record of previous contacts with the competitor companies. These data are collected by the appointed contact centre and organized into a ‘Data Warehouse’.

The content of data file will be customized to the specific needs of the company and inferences will then be drawn about an individual customer’s need.

Data Warehouse and Data Mining in Insurance Business

Data Warehouse and Data Mining is the secret of insurance selling success. This must be thoroughly understood by those who are in the business of selling insurance. Each customer is a subject of record in the data base that provides the customer’s or the prospective customer’s personal details and other contract information. Companies try to capture information from a customer every time he comes in contact with any of its issue focus other hand, extracts information from a database, the existence of which is not known by the users.

Data mining unearths the relationships between variables and customer behaviours that are non-intuitive. Extracting hidden patterns of customer behaviour can help in finding an altogether different route to solve a business problem. Data base of the customer information when combined with sophisticated analytical techniques makes possible to derive substantially precise information on customer needs and trends and monitor changes and variations over a period of time.

Through application of Data mining software, companies can predict future trends and behaviours, thereby allowing business to gather hidden predictive information about their customers to solve business problems.

Whether it is to increase market share or improve internal productivity or gain a competitive edge, data mining is the solution to most problems and issues. We see that today all big companies have made data mining an integral and continuous part of their business processes as database management helps in the process of building, maintaining and utilizing the databases on the customer for the purpose of contracting transacting and building relationships.

For example, consider a life insurance major who wants to launch a new policy and needs to decide on his target customers for offering this particular product.

There exists a historical database consisting of age, address, profession, income, previous insurance coverage, etc. of all existing and prospective customers with prior interactions and responses. The data mining software would use this historical information to build a model of customer behaviour that could be used to predict which customers would be likely to respond to the new product, launched.

By using this information, a Marketing Manager can select only those customers who are most likely to respond, thereby saving time and being cost effective.

Data Mining and Relationship Building

Data mining has acquired immense significance in the contemporary insurance business relationship building. The concept of Customer Relationship Management, involves use of knowledge and analyses about customers with a view to effectively sell them more and more goods and services, and facilitate enhanced customer satisfaction. It is the CRM function,. Measuring customer profitability requires data that relate to both revenue generation as well as the costs of serving the product or service.

It is a fact that in the organization that ushers in improvements in customer service to facilitate long term sustained customer satisfaction and paves the way for repeat purchase, improved customer loyalty, reduced customer switch over and of course greater profit and revenue for the firm.

Virtually every company knows that 80% of its revenues are coming from 20% of its customers. In insurance industry, with high policy lapsation, it is to be found on an average over 45% of customers as unprofitable. Measuring customer profitability requires data that relate to both revenue generation as well as the costs of serving the product or service.

Imagine that you are the marketing manager for an insurance company. You are responsible for managing the relationship with the company’s existing customers. One of your immediate concerns is churning of business and maximizing customer retention.

In this business of insurance you understand that the cost of keeping existing customers around is significantly less than the cost of getting new customers, so you need to figure out a cost-effective way of managing this problem.

Traditional Approach – Pick out your high end customers (that is, the ones who pay heavy premiums) and use your persuasive skills to encourage them to pay their next premiums on time and reduce lapsation and loss of contract. You may offer a rebate on late payment charges, or a New Year gift, but this solution is probably very wasteful. These high end customers would be willing to stick around without receiving a persuasion or a gift.

Modern Approach –The solution to the churn problem has to be perceived in an altogether different way- the customers to concentrate on are the ones that will be leaving! There is no need to worry about the ones who will stay. Instead of providing the customer with something that is proportional to ‘Their Value’ to your company, you should instead provide the customer with something proportional to ‘Your Value’ to them.

All customers are different and they need to be understood individually in order to optimize relationships. High value customer might value the relationship because of your trust and reliability, and thus wouldn’t need any persuasion to stay on. While on the other hand, small to buy new products. They may also recommend the insurer to friends and family members,. In addition, an insurer with a reputation for effective customer service increases its ability to attract new customers and to recruit and retain its sales force.

Improving profitability

Data mining can help in retaining existing business and can greatly increase an insurer’s profit margin, because it is less costly to keep an existing customer than to gain new customers. Avoiding the time and resources required to replace customers and agents lost due to poor services allows the insurer to direct those resources into building its business.

Organizing of Customer Service with the help of Database

Customer service departments can be organized according to the types of business an insurer does and according to the approaches of each insurer’s management. Some insurers establish one department or area to deal with every kind of service to its customers.

Other insurers divide customer service activities by product, by distribution system, by client, or by type of service requested. For example, in organizing customer service activities by product, an insurer usually trains specific employees to respond to the questions concerning one or two products – for example, in case of key man insurance or a group life insurance product, an employee who understands all the details of the product, is the one to answer customer queries.

If the Data mining results show higher number of internal customers, i.e. more agents, sales managers, the department can be grouped in a manner, that so the  internal and external customers are handled separately.

If the results of Data mining show that there are requests for multiple services, assigning employees to different types of requests for service will be the best alternative. Using this organizational mode, an insurer can assign certain employees to make minor changes, while others to work with policy loans, assignments and so on.

It has been seen and experienced that ‘observed’ purchasing behaviours are more powerful predictors of future buying behaviours. It has been seen that the customers who are financially constrained but need the coverage of life most will require your personalized care and concern to stick on with you. The key is determining which type of customer you’re dealing with. The approach of building relationship will be optimized through extracting hidden information.

Applications of a Database in CRM functions

Efficient customer service enhances the insurer’s reputation for taking good care of its customers and in a way creates a relationship bonding which helps and assists agents in making further sales. We understand that customer service is that crucial part of insurance administration that helps maintain important contractual and business connections between the insurer and its customers.

There is an astute relationship between customer service and policy conservation. An insurer that establishes and maintains superior customer service enhances business in a number of ways. Data mining is one part of a much larger series of steps that takes place between a company and its customers.

The way in which data mining impacts the relationship building process depends on the total organizations’ vision and mission and not only the data mining process. The customer database can however have broad range of applications.

Meeting Customer Requirements

Database approach to relationship building employs ‘observation’ rather than ‘inference’ about customers’ needs and behaviour. It has been seen and experienced that ‘observed’ purchasing behaviours are more powerful predictors of future buying behaviours.

Building long-term customer loyalty

By customizing the service approach, insurers build upon the brand loyalty thereby creating such customers who are satisfied with an insurer’s service and are more likely to renew and increase their current coverage and issue focus

Service Improvements for Agents through Data Mining

Agents need information in the form of organized data to service their existing customers. Policyholders consider the agent to be their link to the insurance company and, therefore, communicate directly with him when they have questions or issues. Answering customer’s questions promptly, is vital to a continuing relationship between the policyholder, the agent, and the insurer. If the insurer makes errors in transactions, is unresponsive to requests for information, or otherwise fails to meet policy owner’s expectations, the agent is the one to face their wrath. Providing reliable and effective service by the organization also helps in retention of successful agents to sell an insurer’s products.

Conservation of Business and Improving Business Persistency.

To get premiums regularly and timely is what every insurer seeks. Conservation is the process of ensuring that the insurer retains policies in its books for the entire term of the plan.

For many insurers, conservation of existing business is as important as selling new policies. Data mining helps in predicting and forecasting the lapsation behaviour of the clients, and it is here that trends can be captured in time and remedial measures taken before much harm is done to the business.

Creating Customer Bonding

A strong relationship between an insurer and its policy owners ensures automatic growth in business. This relationship can be built and enhanced through improved customer interactions.

Data captured through customer enquiries, alterations in policies, complaints, and other services related issues enables the insurer to do an in depth analysis to enhance its relationship with its policy owners. By prudent data analyses, an insurer can often identify and help concerned policyholders not only achieve satisfaction but strive towards ‘customer delight’.

Conservation of Orphan Policyholders.

Since agent is the prime link between the insurer and the client, without regular communication and touch of the agent, a policy owner could be likely to cancel or surrender the policy and purchase coverage from another insurer whose agent can provide ongoing service.

Through Data mining, some insurers have been able to establish electronic tracking of agent activities to identify agents who are no longer active. The sooner an insurer can identify an agent who wants to loosen his ties with the organisation, the more likely the insurer can conserve that business.

To Control Claim Ratios and Fraud Detection

Data mining can be effectively applied to detection of fraudulent claims and fraud analysis in various ways. While analyzing the characteristics of fraudulent insurance claims, prior data like type of insurance, vehicle type, age of policy, age of insured, postal address of insured shall be used for detecting hidden information.

Here a Policyholders consider the agent to be their link to the insurance company and, therefore, communicate directly with him when they have questions or issues.

Model might be induced that shows a high propensity for fraud among motorcycle vehicle owners below the age of 25 for policies that have been in force for less than three months. Another approach involving clustering and algorithms could also be used to detect the cluster most worthy of further analysis by reviewing relationship among attributes.

To ensure that customer service is prompt, courteous, complete, and accurate; an insurer from time to time needs to amend its ways and methods and ensure effectiveness in dealings. Providing complete service is sometimes a challenge, especially, to cover a variety of transactions or when transactions are complicated or unusual.

Thus, ‘replace older technology’ is required. The system, in which that a company invests must have the ability to handle future predictive models and applications; and allow to increase market share, internal productivity etc, and gain competitive advantage.

Data Warehousing, is highly relevant and critical to insurers, combines the data from multiple and usually varied sources into one comprehensive and easily manipulated database.

By : Dr. Ashish Barua, Proffesor , National Law School, Jodhpur, Published in The Insurance Times, June, 2011