Customer intelligence initiatives hinge on the ability for the data warehouse or data mart to continue to provide reliable, clean, integrated information. As the database encounters problems and quality deteriorates, user confidence starts to erode, which is the beginning of the end for customer intelligence.

Production support is ever the afterthought for most development projects, regardless of how much production support tries to be involved early and regardless of how many processes or documents are in place for development teams to complete before they go live. A poorly planned production support strategy leads to technical and user adoption issues. To counteract these tendencies, organizations can leverage the following concepts: data stewardship, dependency matrices and proactive data quality assessments.

Data stewardship has been talked about for a quite a while, but many organizations still have not started the first tasks in order to build a data quality organization. One of the reasons for the lack of data stewardship is the daunting task that the literature implies. We have found that there are really three main components that can make a data stewardship program successful:

Ownership. Simply put, somebody must feel that it is his/her responsibility to keep certain subject areas of information clean and reliable. Though it is preferred that stewards be business owners, IT can also play the role.

Analyst Support. Most data owners do not have all of the technical ability or technical knowledge to research quality issues. Consequently, stewards need part-time resources that can investigate, trace and document the information in order to understand if quality issues are system bugs, process issues or business rule issues.

Some Authority. Data stewards do not need to be supreme dictators; however, as issues arise, the ability to influence the priorities of source applications and processes may be the key to resolve many issues.

Other data steward initiatives, such as formal ongoing quality improvement processes, enterprise-wide meta data repositories, incentive and bonus structures, quality dashboards and metrics, and weekly status meeting are important, but should not be forced from day one.

Dependency matrices help production support and source system applications understand where the data is coming from and where the origin of a problem may arise. Dependency matrices typically outline: source applications and tables for the customer intelligence application, frequency of data transfers, individuals responsible at the source for the information and the data transfer, and target location for the information.

Dependency matrices have the bidirectional impact of making sure source systems understand the importance of the integrity of their information on dependent applications as well as helping customer intelligence production support organizations understand the lineage of the information. Some extract, transform and load (ETL) development tools automatically generate dependency matrices.

Proactive Quality Assessments - Automating the detection and reporting of data quality information is the most beneficial to a production support organization. Scripts, programs, error files and reports help automate what many organizations call data watching. Data watching provides the ability for customer intelligence support organizations to check the health of the system's data integrity each day (or week or month). These quality checks allow support to catch issues before users even know there is a problem. Data watching typically catches: source system changes that have not been communicated to the customer intelligence team, data that has not complied with business rules, numerical or financial checks and missing or duplicate information.

Production support is not sexy; however, it has become critical to the success of customer intelligence. The front line of support that answers user questions, investigates data issues and coaches data stewards regarding possible resolution are the heroes of customer intelligence on a day-to-day basis. Appropriate focus on support has been as critical as the innovation from the analysis of the information to reaching customer intelligence ROI.

Register or login for access to this item and much more

All Information Management content is archived after seven days.

Community members receive:
  • All recent and archived articles
  • Conference offers and updates
  • A full menu of enewsletter options
  • Web seminars, white papers, ebooks

Don't have an account? Register for Free Unlimited Access