Governing Customer Information in Insurance

Customers and insurers are now starting to interact with each other using a wide variety of different channels including agents, call centers, web, SMS, email, and social media. This is increasing the need to integrate these channels and manage the data and contacts accordingly. It also raises the question about who owns the customer. Many insurers have taken the view that nobody owns the customer, but that many people may have different relationships with the customer. This requires strong information governance to determine who is primarily responsible for certain attributes of customer data. There are various types of insurance products including life, property, automobile, and specialty. Each line requires specific customer information depending on its unique needs. Although the specific customer information will vary by line, we have included some examples of key attributes and the associated information governance challenges:

  1. Identity information

Information such as name, social security number, date of birth, and party type is critical to link multiple policies with the same customer. For example, life insurers typically issue policies over an extended period, and it is important for the underwriting department to quantify the overall exposure to any given individual. In another example, a major insurer stored the records for all parties in the PERSON table irrespective of whether they were persons or organizations. This made it more difficult to apply person matching rules in all cases. The information governance program moved all organizations to an ORG table so that it could apply matching rules that were more tailored towards organizations. Another insurer ignored the suffix such as “Jr.” (junior) or “Sr.” (senior), which created inaccuracies in matching policyholders. Finally, a large multi-line insurer found that more than 200,000 parties had been bulk loaded into the customer information file (CIF) without an active policy. This made it hard to implement cross-sell and up-sell programs.

Case Study: Linking of jumbo life insurance policies

A large underwriter of jumbo life insurance policies wanted to quantify the total exposure to key individuals and households. The overall sponsor for customer information governance was the underwriting department. Customer information governance was difficult because the life insurer would issue policies to the same individual or household at different times. It was not easy to match policies because an individual might buy a life insurance policy today, and then buy another policy several years later. During the intervening period, the individual’s address might have changed, as would their name due to marriage or divorce.

The life insurer set information governance policies that dictated that IT would use data quality tools to conduct the initial policyholder match across individuals and households. The match rules were not based solely on United States social security numbers for two reasons. First, the team was concerned about the accuracy of social security numbers in their database. Second, the team had a lingering concern about the possibility that the United States Social Security Administration might recycle numbers when individuals die. As a result, IT implemented composite matching rules that included the individual’s first name, last name, street address, date of birth, and social security number. Once IT provided an initial set of match candidates, the policy administration department completed the final match.

  • Date of birth
    Although date of birth is part of identity information, it is an important attribute in its own right because it is used to calculate premiums especially for life cover. Let us review a case study from a European financial services conglomerate.

Case Study: Information governance around date of birth

A European financial services conglomerate owned both a bank and a life insurance company, and was in the process of implementing a centralized master data management (MDM) program to better manage customer relationships.

The information governance team reached agreement on the 21 attributes for “customer” data. The information governance policy was that a customer could make a change to his or her profile by calling either the bank or the life insurance company, and have those changes cascade across the entire enterprise. This process created some information governance issues. The bank would accept changes to date of birth with limited documentation, while the life insurance division would require supporting documentation, due to the impact on life insurance premiums. The bank believed that the life insurance process would slow them down, and the life insurance company felt that the bank’s processes were too risky.

After much hand wringing and several internal meetings, the information governance council brokered a compromise. The customer service representatives in the bank would have a screen that flagged customers who also had a life insurance policy. If the customer did not have a life insurance policy, the bank would accept changes to the date of birth without documentation. If the bank customer also had a life insurance policy, the bank required the customer to submit supporting documentation.

  • Contact information
    This includes phone number, email address, and mailing address. Marketing, telesales, and customer service may be sponsors for information governance around contact information. Marketing needs good contact information to support campaigns. In at least one known instance, a life insurer called the same person 10 times during the same weekend, as part of a telemarketing campaign. It turned out that the agent had entered his own phone number instead of the customer’s phone number, and ended up receiving multiple calls. In another instance, the information governance program at an insurer found several instances of multiple emails for the same party. This was despite a stated policy that there was to be only a single email address per party. In yet another instance, the telesales department found that an improvement in the accuracy of work and mobile phone numbers increased sales per hour by eight percent. Finally, work address is an attribute that is rarely maintained by life insurers. The unfortunate circumstances surrounding the September 11 attacks highlighted the importance of this attribute. Several reinsurers contacted their life insurers to quantify their total exposure to individuals who worked in the World Trade Center in lower Manhattan. Obviously, the life insurers did not have this information readily available, and the reinsurers had no way to quantify their exposure in an expeditious manner.
  • Gender

    Gender is another attribute that drives business decisions at insurers. The European Union recently ruled that insurers cannot provide preferential rates to female drivers. However, gender is still an important attribute for many insurers who charge males higher premiums for life cover. The attached case study provides an example of information governance around gender.Case Study: Information governance around gender at a European insurer

    A European insurer became suspicious that it had might have inaccurate gender data when it received a claim from a customer with a female name but who was classified as a male. On further investigation, the insurer found that the new business department had input gender data incorrectly, which resulted in the customer being billed for incorrect premiums for several years. The insurer then decided to conduct a broader investigation against a reference file of typical names and the associated gender. The investigation discovered several thousand incorrect entries in both directions (males classified as females and females classified as males). While the insurer did not believe that there was any fraud involved, they did realize that premiums had been charged incorrectly over many years for these customers. The insurer informed the regulator who determined that all overpayments by customers had to be refunded, but any underpayments had to be written off. However, the regulator allowed the insurer to write to these customers, inform them of the error, and change future premiums. This situation cost the insurer €4 million. Because of these findings, the insurer changed its new business system to validate names against the reference file mentioned above, and exceptions were highlighted for further investigation.

  • Net worth

    Net worth is also an important attribute that drives segmentation decisions although it can be supplemented with other data as shown in the case study below.

Case Study: Information governance around net worth

A large European life insurer defined high net worth customers based only on the sums insured. The information governance program introduced additional segmentation attributes such as lifetime value and household value to improve the quality of the segmentation analysis. As a result, the high net worth sales team was able to craft more tailored offers to retain these customers.

  • Other demographics

    Insurers may use other demographic attributes within their operational systems and for market segmentation. However, insurers need to be careful about holding transient data such as salary or marital status. At a minimum, these fields need to be date stamped as the circumstances can change and the insurer may not be informed of these changes. More importantly, insurers need to understand why they are holding some attributes (see case study).

Case Study: Information governance around marital status

An insurer was involved in a major internal debate around which department could update the marital status field. Upon further investigation, the insurer found that the data contained a variety of values such as “married,” “single,” “divorced,” and “separated.” However, the field also contained values such as “1,” “0,” “once,” “never,” “yes,” and “no.” On further examination, they found that nobody used the marital status attribute as it could not be trusted, and therefore they decided to dispense with the attribute altogether.

  • Customer profitability
    The retention department at a large insurer sponsored an information governance initiative to ensure the consistent calculation of customer profitability across the business. Several factors including the number of products per customer as well as the length of relationship with the insurer affect customer profitability. Because of the improved visibility into customer profitability, the retention department was able to tailor more targeted customer offers and higher discounts to retain customers that were more profitable.

  • Privacy preferences
    These attributes include customer preference indicators such as do not call, do not email, and do not fax. The privacy department will support strong stewardship around privacy preferences. One of the major benefits of having an information governance program in place is that it involves the business in key rules and decisions relating to data (see case study). Insurers sometimes face a dilemma where a customer has purchased products from two different channels. Consider a situation where a customer has purchased life cover through an agent and then purchases an investment bond directly from the insurer. Because regulators are keen to ensure that the customer receives “best advice,” it is clear that a servicing agent should be provided with the full portfolio of holdings for the customer to advise on appropriate fund choices. Due to agent charters and privacy constraints, insurers may not always be able to provide this information. This can create a key information governance issue that also needs to be carefully vetted.

Case Study: Information governance around direct marketing consents

An insurer was starting down the path of building a single view of their customers. The insurer asked the IT departments from two separate lines of business to work together on the customer hub. Being logically minded, the IT departments established a business rule that said, “if a customer exists in both businesses, but they have agreed to receive marketing offers from one business but not the other, then we will default to the less favorable position (i.e., we will not market to that customer).” IT then merged the customer records into the centralized hub and applied this rule. One of the marketing departments realized what had happened, and immediately reversed the decision. This involved the business taking a risk of upsetting some customers and ultimately being fined by the regulator. However, they viewed this risk as being worthwhile. Lots of work had to take place to undo the changes that had been made by the IT department. Ultimately, everybody felt that the business participation was critical to make the right decision that balanced the associated benefits and risks.

  • Rules of visibility
    When many parties require access to the same customer data, it may be appropriate to limit their views to specific attributes depending on their role. For example, underwriters may require access to medical data that would be restricted to other parties such as customer service.
Let us bring some of these preceding concepts together in the following discussion. The marketing department at a multi-line insurer could act as an overall sponsor for an information governance program focused on a single view of the customer across multiple lines. Multi-line insurers can generate substantial lift by cross-selling policies to existing customers. For example, a customer with a life insurance policy should also have auto coverage and homeowners coverage. In addition, underwriting could leverage this information to establish a single view of the total risk associated with the customer (e.g. the number of policies and the types of policies the customer already has in force with the carrier). Actuarial could use the single view to assess the profitability associated with the multiple products held by a customer, and may look at how the total coverage plays into products per customer and retention. Further, customer service could benefit from operational efficiencies. For example, one large insurer believed that they could create a centralized call center organization and drive economies of scale if they had access to a single view of the customer. In addition, the customer experience would improve because customers would not have to call two different phone numbers at the insurer to update the address on their life insurance and homeowners policies.



submit to reddit