As companies try to differentiate themselves in increasingly competitive markets, their customer relationship management (CRM) systems need `clean` data to enable them to satisfy and retain customers.
Best practices are generally known and are already being broadly applied in most industries. It is the access to, and application of, knowledge of customer and product that really gives a competitive advantage. Establishing a reliable information base must be a primary concern according to Simon Jeggo, GM of Ardent Software Africa, which markets specialist data warehouse tools.
He says: "It is time companies stopped trying to fix `dirty` data and started fixing the underlying causes. The current emphasis on `cleansing` fragmented data has to change. If this is not done, the CRM information is at risk of being corrupted."
The growth of a business culture that leans heavily on customer relationship management (CRM) has highlighted the problems associated with invalid, incomplete and missing information. Information extracted from operational systems is almost always at fault. "If information is not operationally essential it is often just recorded and not checked programmatically. Flags or indicators such as customer type, which may indicate a marketing category, may never be used within the operational system and so never checked", says Jeggo.
"Another complication is that there may be several sources of one piece of information. This is especially true of the so-called legacy systems.
Organisations on the acquisitions trail often suffer as a result of unsuccessful rationalisation of their systems. Names and personal information may be recorded in a number of places, each in their own way."
Efforts are made to rationalise and correct the data and terms such as `scrubbing` are used as data is pulled out of operational systems and loaded into information repositories. But, this situation doesn`t change as the next time the data is extracted; the same errors are present.
Jeggo recommends a permanent solution using a monitoring, remediation and feedback process that corrects the source of the data problems. The approach is multi-pronged: Strengthen the validation rules in the operational system, train personnel to enter real information, not just any default or generic information, get rid of internal redundancies and duplication. All extraction and loading must check and constantly give feedback on the purity of the data.
"Dirty data indicates bad practices and there is a cost associated with this - some companies actually budget for that cost. A simple example is the charge that the post office makes for incorrect postal codes. There is a cost associated with duplicating a customer`s records, with duplication of marketing material and accounts, not to mention the aggravation.
Increasingly, accurate statistics are called for by various statutory bodies that can impose fines for invalid information. Companies that do not have reliable information readily available cannot have an effective CRM initiative and the resulting turnover in customer enrolment is costly in lost opportunities and in the investment necessary to enrol new customers."
Share
Editorial contacts