A silent killer is ripping more than R250m out of SA business each year

That covers wasted time, paper and mailing costs, the cost borne by the South African Post Office (SAPO) for undeliverable items, the cost of lost potential business, late payments, the cost incurred investigating and reprocessing payments, and more - the list goes on.

"Contact data quality boils down to the customer. Customer relationship management (CRM) revolves around one-to-one contact with the individual customer, and if at any stage the direct contact is disrupted, the company will not be able to manage any type of successful relationship with that customer. That's where the focus on data quality starts," says Ardagh.

Companies often cannot quantify data quality - or, rather, the lack of it. They accept up to seven percent of returned or undeliverable mail, and commonly expect it.

"Seven percent may not seem high, but companies need to realise that realistically only a percentage of people who receive mis-addressed mail bother to return it to sender. This means the amount of undeliverable mail is substantially higher than is perceived," Ardagh points out.

The challenge in data quality lies in recognising that data is dynamic, and admitting the quality of your data is not what it can be, explains Ardagh. "An unchanged database is likely to decrease in data quality by two percent each month. This can amount to 24% in a year. If the database is not upgraded and the customer contact information regularly checked, the data of that company will deteriorate significantly over time."

Statistics show that a person will move home every two or three years, with 20-24% of the population moving annually. A common mistake is to assume a customer details remain unchanged simply because a customer's debit order has remained active. A customer may have moved four times without changing his bank details.

Inconsistent information is a common obstacle to achieving quality customer data. For example, an organisation may have separate customer records for Richard Smith and Dick Smith, although both names refer to the same person. Similarly, a company may have two Richard Smiths on file - one on Valley Road, the other on Valley Drive - but both actually the same person.

Things get more complicated, says Ardagh, when the company's CRM efforts want to link people in the same household. If Dick Smith overdraws his cheque account by R100, for example, the bank might return the cheque. But if Dick Smith's father, Richard Smith, has several million rand invested with the same bank, bouncing Dick Smith's cheque could unsettle a very lucrative customer relationship. In this case the bank's CRM applications for managing retail customer accounts and high net-worth accounts need to exchange data accurately.

"Reliable data is crucial for effective CRM. Customer perception is all-important. Why should a customer give business to a company that clearly doesn't have accurate information on its clients? The impression left with the customer is that the company simply doesn't prioritise customer care," explains Ardagh. "If another company approaches the customer with a similar product and an offer of a better service, they may decide to change vendors."

Mailing is not the only area affected by poor customer contact data quality. One dark side of the business case for data warehousing is the failure of operational applications to provide for effective data management of the business-critical information resource. The organisation, warns Ardagh, is paying dearly for this.

Data warehousing projects fail to deliver value for many reasons - one of which can be traced to non-quality. Poor data architecture, inconsistently defined departmental data, an inability to relate data from different data sources, missing and inaccurate data values, inconsistent uses of data fields and unacceptable query performance are all components of poor-quality data and reflect directly on customer contactability. To illustrate, a perfectly designed database can only provide services based on the lowest common denominator of data quality.

"A data warehouse has multiple feeds from different sources throughout the organisation. Most companies have no idea how many duplicated records they're laden with, or mis-spellings they carry in their database. They also have no idea what that's costing them in terms of wasted transactional time, poor time management in their data warehouse structure and, importantly, perception cost - the perception the customer gets when a company gets their details wrong," says Ardagh.

A critical success factor in CRM initiatives is accurate customer contact data. Data needs to be cleansed, initially and on an ongoing basis, for business to be assured of a reliable information platform. Without accurate customer data, there might be no customers; without customers, there can be no business.