Master your data and you’ll master omnichannel

Tuesday, Jun 28th 2022
Hand holding smartphone, interconnected, using data, and other hand on computer, omnichannel

Data is collected and processed in a variety of systems; CRM and financial applications, call centres, external data suppliers and more. Every data source has its own unique format, which means that customer data is stored differently depending on the particular system, and probably exported in a variety of ways as well. The syntactical and semantic validity of data sources varies greatly, such that a particular individual stored on multiple systems may not be recognized from one to the next. In addition to inconsistency, data may be incomplete, inaccurate, or just plain outdated. 

Poor data quality causes innumerable problems that will continuously harm your omnichannel communications. Consider these examples: 

  • A standard address field in the CRM system limits the number of characters, so customer service frequently enters all or part of the address in the field for Other. Subsequently, 22% of a mailing is returned for missing or incomplete addresses.
  • Clara Pierce used multiple email addresses with a particular vendor; her full name, clarapierce@, her last name, pierce@, and the joint email address she uses with her partner, claraandchris@. As a result, she received three emails for a single promotion. Assuming that she is not the only customer with multiple email addresses, the actual reach of that marketing campaign was much lower than expected.
  • New customers are sent a promotional offer for travel insurance. Because transaction dates were not registered in the marketing system, the notice also goes to existing customers who had recently paid full price for the same insurance. Not the optimal customer experience.

In theory, all those problems would go away if you had a team of employees with training in pattern recognition, context analysis and other linguistic interpretation skills. They would recognize that Aaron & Peter Whitham Logistics, Colegate 7, Norwich and A. Whitham & Son, Transportation Mgmt., Colegate, Norwich were the same. But your systems see these as two completely different records. 

Fortunately, there is intelligent software capable of natural language processing that can ensure that your data quality is top-notch. It simulates human reasoning so it knows that Elizabeth is typically a female name and will change Mr. Elizabeth Peters accordingly

Typically, intelligent data quality software will allow you to perform the following activities: 

• Validation

• De-duplication 

• Cleansing

• Standardization 

• Capitalization

• Enrichment 

Data is both complex and volatile; it is collected quickly and may become outdated just as fast. For instance, people move, so some portion of registered addresses is obsolete at any given time. Even if the address is correct, the formatting may be wrong, like a name whose initial letter is not capitalized. And, as in an earlier example, it is likely that customers are registered in the database more than once. You can see why it is important to revisit the state of your data on an ongoing basis when you want to build a solid omnichannel strategy.

These are just a few examples of how data quality influences the quality of your omnichannel communications. To explore this topic further, and learn what other factors might be of influence, we invite you to download this free whitepaper: “The recipe for success: Consistent customer data and an omnichannel strategy”.