Update on Customer Information Quality Discussion Group

Dear OASIS Members,

This message comes to you from Ram Kumar (rkumar@msi.com.au) of Cognito, Inc., an OASIS member. Ram is the Discussion Group Leader for the Customer Information Quality discussion group. Please read further for detailed information on the intent and scope of this group. If you should have any questions, or if you should wish to join the group (ciq@lists.oasis-open.org), please contact Ram directly.


The objective of OASIS Customer Information Quality (CIQ) discussion group is to initiate an OASIS Technical Committee to establish and promote a uniform, unique and identifiable customer profile, carrying elements to find and reach the person/company behind the profile.


Research tells us that it is much more cost effective to retain and invest in our existing customer base than to build or buy market share. One customer lost through ineffective marketing means a massive expenditure on acquiring new customers to make up the lost revenue. This is more apparent as we move into the e-business environment. As companies move to establish more effective relationships with their customers, the need to achieve a complete view of each customer's dealings is recognised as critical.

A direct result of this is the development of company-wide customer relationship management strategies, representing a combination of business processes, information management tools and importantly, customer data. Customer data forms the foundation to build effective customer relationships. To be effective, customer data must meet the highest possible standards of both quality and integrity.

Often, it is only when attempting to unify customer data from disparat ebusiness systems that the impact of poor quality customer information is fully understood. While data within individual databases may be fit for the purpose for which it is collected and used, combining data from a range of sources for a new and different purpose poses a real threat to the effectiveness of entire customer relationship initiatives. The bottom line is that reliable and accurate customer information is now more than ever essential in establishing effective customer relationships and therefore, customer information quality management is critical.

The Internet threatens to turn customer marketing on its head. While it has created tremendous opportunities for relating to customers on a one-to-one basis, it has also created a nightmarish challenge for companies struggling to understand this new way of interacting with customers.

Despite the tantalising opportunity it represents, e-business is still an inefficient way to attract and retain customers. Customer acquisition costs on the Internet have skyrocketed to US$65-250 per customer. Churn is up because more than 50% of Internet companies cannot respond to their customers. Despite the promise of the Internet, companies do not know and cannot relate to their customers due to the poor quality customer data they maintain. To make matters worse, Internet customers are a global audience. Global customers require 24-hour-a-day, 7-day-a-week service, have language differences, multiple data and shipping formats, and vast differences in demographics, tastes, preferences and so on.

The general belief is that the more we know and understand about each customer the higher the chances they will purchase and return.Therefore, retaining e-business customers by maintaining reliable and accurate information about them is now the most critical problem in an e-business environment. Customer information quality management is the answer to this problem.


Customer data consists of many components. However, a person or company's name and address is the key identifier of a "customer". Name and address, as a data type, is very difficult to manage. This data is often volatile... customers come and go, addresses change, names change. This data is often cluttered when entered. Name and address fields on data entry screens are usually free format and ripe for users to enter comments without any edits. Name and address is subjective...it can be written in a number of different ways and still be the same. There is no application independent standard to represent name and address data and to measure its quality. This problem is further compounded by the different ethnic backgrounds of name and address data in a global market. An application independent XML standard for name and address data management called Name and Address Markup Language (NAML) has been developed and this will be discussed and further developed by the CIQ committee. Though NAML does not include all the address components throughout the world, it is a good start. NAML will evolve as more additional name and address components are identified. Businesses dealing with name and address data (eg. mailing houses, address verification and management services, name and address data quality vendors, etc) can use this standard for managing name and address data.


Although name and address data is the key identifier of a customer, other data helps to uniquely identify a customer. Customer addresses frequently change and it is not trivial to link the customer across multiple addresses with just name information. In the example below, a customer can have two completely different addresses and it is nearly impossible to uniquely identify the customer with the name alone. Customer centric data such as telephone numbers, e-mail addresses, account numbers, credit card numbers etc. will be necessary to achieve this. It is therefore, important to define a standard for representing all forms of customer data that forms the basis of customer information quality management. An application independent XML standard for customer information quality called Customer Identity Markup Language (CIML) has been developed and this will be discussed and further developed by the CIQ committee. CIML provides a framework for representing different data about a customer and as a result, helps to uniquely identify a customer. Given that name and address is a subset of customer data, NAML is a subset of CIML. Businesses dealing with customer data can use this standard.