Most loyalty marketers have real concerns about the quality of their data, according to Greg Council, vice president of marketing and product management, Parascript.
Loyalty360 talked to Council about this critical piece of the customer loyalty puzzle.
What are the challenges for loyalty marketers as they relate to creating high-quality customer data?
Council: The Harvard Business Review reported recently that the cost of poor quality data in the United States for 2016 alone was over $3 trillion. This averages out to costing mid-size and large companies millions of dollars per year. Data quality challenges hit the big three: It’s time-consuming to fix; it’s expensive in terms of lost or missed opportunities; and it results in reduced customer satisfaction and loyalty. Unreliable data usually means the staff is checking the data to find any errors and make corrections. Later, the business has to deal with the fallout from those errors that inevitably slip through the process and negatively impact the customer.
Given these challenges, what can be done to improve customer data for loyalty marketing?
Council: Data quality may be the single largest hidden factor impacting loyalty marketing. While there’s no “easy button” to get high-quality data, there are ways to ensure quality:
Treat data quality as an ongoing, moving target. Like SEO and your website optimization, there’s no one-time fix. It requires attention.
Standardize your data management processes across the company. Data is usually coming in from many different sources across multiple departments and functions so establishing a consistent data collection process is important.
Start at the top to educate leadership about the current data challenges to make the necessary investments to overcome any long-standing data management challenges and drive increased market performance.
Adopt and leverage the right technologies and tools that help ensure quality data is captured and eliminate incomplete, out-of-date or duplicate data.
How does a brand know if it has refined, high-quality data?
Council: To really know you have quality data, you have to test it. An enterprise-wide process for automatically validating incoming data at the point of capture helps ensure quality data from the get-go. For instance, address data can be immediately validated against a postal database, and there are numerous other third-party services that can act as similar references for needed information. Continuous data-cleansing, validation against the data you know is correct can be an automated process.
Constant refinement to these automated processes will help ensure you have high-quality actionable data that also complies with security and privacy regulations. With effective validation processes in place, either automated or manual, you know when there’s a data issue. And here’s another thought–why not let your customers see the data you have on them? There is a mutual benefit with good quality relevant data and customers will often help to correct key data.
Can you talk about “trapped” data and finding a trusted data source?
Council: Trapped data is, typically, data that’s hard to get at because it’s in an unstructured format. Trapped data is a lot easier to access than in the past using new technology. This data can help reveal an important customer journey or breakdown in service with consequences like an error in double-invoicing when examined together with customer service interactions, a consumer feedback survey that ultimately ends with a closed account. Extracting trapped customer data from documents that come in all types of formats and delivering it in a digestible format shared across the enterprise allows marketers to use their data to stay on top of customer satisfaction and grow loyalty.
What trends do you see as far as leveraging high-quality customer data this year and beyond?
Council: Let’s look at analytics applied to a very common document–the receipt. Many billions of receipts are processed each year. A lot of this data is entered manually although it can easily be automated—even fields like handwritten tip amounts and totals. So, once this data is located, extracted and verified, it can be used in predictive analytics.
This is the next level of service that exists and promises to have significant positive impact upon the B2C markets and the customer. Applying pattern recognition to receipt data identifies hidden patterns of purchasing—spending and non-spending. It can identify the likelihood of type of spend, level of spend where and when along with other demographic data.
Predictive analytics capabilities can be applied to the output data streams of today’s receipt processing services provided to companies. This is today’s progression towards consumer-centric, personalized service and targeted advertising without increased costs or burdens on businesses.
Once marketers can aggregate expenditure data with a high degree of accuracy, probably one of the biggest areas of interest (and concern) is the increased application of psychographics on customer created information such as social media to derive potential interests and motivations. See www.applymagicsauce.com for examples on how it works.