We all know that ‘bad data’ is bad, but to what extent? Have you ever been a part of an email campaign where you get more email bounce-backs than successfully sent messages? While the annoyance factor is high, research tells us it can’t compare to the true cost of having customer data that’s out of date, duplicated, or inaccurate. As a frame of reference, think about the grocery store loyalty cards in your wallet. Do you always have them with you when you
According to The Data Warehouse Institute, inaccurate and low-quality data costs U.S. businesses $611 billion each year. This takes into account bad mailings and staff overhead but does not necessarily account for all the soft costs of having bad data, which may have an even larger impact. For instance, consider the implications of frustrating or alienating a loyal customer by misidentifying them through a website login. Or what if your major airline fails to recognize your rewards program status when you try to book a reward ticket or take refuge in the airline’s lounge. Yikes! The bad data problem is that serious. Losing a loyal customer has repercussions that go beyond a single bad experience. That organization may lose future sales, possible referrals, and be at risk for negative comments on social media−all from a once-loyal customer.
Bad data is out there, but companies are making effort to minimize it. According to research by Jigsaw, the market for data integration, management, and hygiene is $6 billion. And, more than $10 billion annually is spent buying, managing, and cleaning data. If your organization can ensure that quality data is stored in your data warehouse or business intelligence application, you can also ensure the quality of information for dependent applications and analytics. Effective organizations understand that quality, or ‘golden’, customer data is the lifeblood of their CRM systems, not to mention overall business success. We define ‘golden customer data’ as accurate, validated, and consolidated customer records that empower the company to provide a better overall customer experience. These records demonstrate the true lifetime spending and value of that customer and indicate their preferences, tastes, and likelihood to buy again. Companies with access to this caliber of customer data can foster stronger customer connections and stronger ROI of BI and CRM systems.
WHEN GOOD DATA GOES BAD
It’s not hard to imagine how easily good data can go bad. Do you ever get two catalogs in the mail from the same business? If your organization has two or more records for just 25% of your customers, that report showing the count of customers could be off by 25%. No wonder the sales team didn’t meet the goal last month!
In another example, we see customer data where there are more than 10 different entries for something simple, like state names (i.e. Colorado, Colo, CO. Co., etc.). This adds unnecessary report development time. We’ve also seen organizations with product data that was duplicated and given a new part number for every change to a label, causing countless inventory and shipping issues.
WHAT MAKES UP YOUR CUSTOMER'S DATA?
Whether your source is the customer, vendor, product, or employee, working towards gathering golden customer data does not have to be an overwhelming project. However, it will be an ongoing project. The moment data is collected, sorted, and stored, it starts to become stale. The job of data advocates is to have a master data quality management plan that puts processes in place to continually capture accurate and complete customer information.
WHIP DATA INTO SHAPE
While it’s a challenging endeavor, an organization can take steps to cleanse data. The first step is to create data advocates within the organization. The more executive-level attention data quality receives in your company, from the CIO and up, the better your data management program will be. It’s not enough for executives to hear secondhand about the difficulty of creating reports. Business groups may spend countless hours each week creating meaningful reports, but executives rarely feel this pain. Groups need to capture the productivity drain in dollars and cents when dealing with data quality issues. Other important considerations:
- Profile your data - At the start of any data quality management project, it’s important to profile your data. Data profiling analyzes the content of critical data items and tests whether the data is fit for your business purposes. Conduct data profiling and measure gaps
againstspecified business rules.
- Scorecard data - Prepare a data quality scorecard. A summary chart will give a business-wide overview of the data quality and show a view of the data quality score against targets.
- Kickoff a master data quality management initiative - It’s important to create a single version of the truth for one data set at a time. In your data management program, set up rules for defining customers, such as your customer’s last purchase date. If the time since the last purchase is more than 1 year, ask yourself if they should be in your database.
Also, create rules to standardize your data. Develop state standards, country ISO, phone numbers, etc. This requires conversations between business groups, from executives to business report writers to developers. For example, during application development, it’s important for developers to create online forms in sync with reporting requirements. When a customer says she lives at 123 Blake St., the system automatically adds the street description, such as 'street.' This ensures the customer record is complete. This small change has a huge compound effect. It reduces call center time, the time it takes to correct addresses, lost mailings, and other costs. Consider using a service provider to tackle a small portion of your data-quality program first.
- Create a competitive data quality environment- Create a competitive atmosphere between sales, call centers, marketing, accounting, and other teams, with metrics to show data quality improvements. It’s proven that data quality can be fun and rewarding for an organization. This builds a sense of teamwork and has a snowball effect. Once the master data quality management plan is in motion and teams are invested, new data captured will be consolidated, cleansed, and lead towards a golden customer record. This is a cyclical, iterative process that requires prioritizing critical issues first and broadening the scope as you resolve these issues. Data Quality is an ongoing discipline, unlike a one-off project scope.
Understanding that high-quality data is critical for effective customer communication, connection, and accurate customer analysis is just one part of the equation. The other part is having the commitment from data advocates to implement an ongoing data quality initiative. Taking steps to ensure data quality will get you closer to capturing golden customer data and provide a better customer experience, improve customer connections, and reduce costs.