While companies often deploy customer data in a wide variety of different areas, ranging from human resources to sales and everywhere in between, everyone should be more or less on the same page when it comes to data quality. All professionals, regardless of their specific line of work, should have an interest in accurate data - in HR, it helps you know your employees better, and in sales, it empowers you to deliver the perfect sales pitch.
Establishing data quality, though, is often harder than it looks. One inherent problem is that everyone within a given company needs to agree upon standards for quality. How accurate is accurate enough? What specifically are people looking to do with their data, and what level of quality do they need in order to make it happen?
Negotiating the right parameters
Different departments are going to have different needs when it comes to data quality. A marketer might need to have accurate contact information for certain customers, while customer service reps might instead focus on having knowledge of people's communication habits.
According to Waters Technology, the key is to negotiate the right data quality parameters so that everyone can get what they need. Patricia Huff, manager of global client data services at the Royal Bank of Canada, admits that this is a complicated process.
"Not everybody's needs are met, but the vast majority of them are," Huff said. "At that point, does the downstream consumer have to maintain something separately, or is the next phase of negotiations about saying they need additional data in the system?"
Figuring out the right data quality standards is an ongoing process. It's dynamic, based on the changing needs of each individual person that relies upon customer and employee data.
Can everyone walk away satisfied?
Of course, the ultimate goal when establishing data quality is to make the masses happy. You want to negotiate standards to be so comprehensive that everyone walks away satisfied. Michael Shashoua, editor of Inside Reference Data, is concerned about whether this is possible.
"A question remains," Shashoua recently wrote. "If neither data managers nor business operations managers are likely to walk away satisfied from a negotiation on how to proceed with a data quality effort, as Huff suggests, is it really possible to get meaningful improvement to data quality?"
It's a difficult question to answer because "satisfaction," like "data quality," is a murky term with a subjective definition. Companies may never fully grasp what it means to have satisfactorily high-quality data, but they'll need to keep working to fine-tune their approach.
