Make No Mistake: Avoiding Data Governance Issues

Rob: The biggest data governance issue is not having a data governance strategy or plan for your organization. Many organizations ignore or avoid data governance for any number of reasons. The most prevalent of these reasons we encounter is the difficulty in demonstrating the business value of data governance to the organization balanced with the cost and time commitment. This obstacle can be overcome by documenting and collating instances of data issues and the resulting impacts to make a solid case for a comprehensive data governance program. Like any program, ongoing management, review, and audits are required to grow the strategy and align to organizational objectives and ever-changing laws.

Heidi: Another data governance issue we see is perpetuating the problem. As companies implement new systems, they may simply replicate what existed without the necessary data cleansing. Yes, it is painful, and yes it can be time-consuming, but upfront data cleansing and consistently adhering to data governance processes is time well spent to eliminate issues downstream.

Heidi: Data governance, including product or location attribution, can help online consumers—as well as associates—get to your products and spaces faster. If I were looking for a glass pumpkin for my home, and the retailer did not tag it as “glass,” it may not show up in search results. Similarly, many retailers offer different assortments by location (e.g., no pet toys in Whitehall). To provide a consumer with the best knowledge of what is available (both online and in-store), these tags can be the difference between a profitable transaction or a lost sale. As another example, in assessing inventory risk after the winter holiday season, products must be consistently identified. If half the organization uses “Holiday” and the other uses “Christmas,” it will be difficult to have a full picture of inventory liability in your reporting.

Heidi: Here’s a real-life example we still often see in client situations. When I was the head of retail planning for a luxury retailer a few years ago, our CEO would conduct regular store visits. We would produce a book for each store, for him to prepare for the visit. It took hours to produce each book because there were multiple data sources, inconsistent attributes, and different spellings for store names in each system. The CEO was oblivious to the intense manual effort required to produce the books because we would do whatever it took to get the job done. A formal data governance process would have avoided much of the manual effort.

Rob: Additionally, effective data governance ensures policies and procedures are in place to reduce errors and block potential misuse of data. When implemented and executed appropriately, data governance ensures all data is used properly and sensitive data is protected in compliance with data privacy and protection laws.

During a recent implementation of a retailer’s global merchandising system, it quickly became obvious the client did not have a solid data governance strategy. North American and European operations used completely different master data hierarchies. This made the consolidation and loading of history into a single global platform unnecessarily more time-consuming and difficult. In addition, other downstream systems like demand and order fulfillment that rely upon clean historical data were unable to achieve the stated ROI goals and objectives because of the inaccurate data.

Heidi: Have a strategy and have a data governance leader. This leader is someone sitting at the table with other senior leaders. They must not only determine the rules to keep company data clean and protected but also lead and promote the company’s overall adoption of data governance processes and policies. This role should focus on business needs and be a bridge between supply chain, merchandising, planning, product development (amongst others), and IT. Business resources define “what” is needed, whereby IT figures out the “how” to meet those needs.

Rob: Start thinking about data governance as early as possible, if not for the entire enterprise at least for your business unit or department. It is never too late to begin a data governance program. There are many great articles and resources on data governance through respected advising groups like Gartner. Of course, you can always reach out to The Parker Avery Group—our team has implemented hundreds of systems for retailers, and we understand the data challenges inherent in the retail industry.

Rob: Data governance is a big topic that encompasses several areas, including your company’s ability to efficiently operate and serve your customers. The effort to implement such a program may seem great, but the ramifications of data issues or breaches can permanently damage your organization’s credibility.

Heidi: The way a company structures and organizes its products and locations is crucial to providing consumers with information about where to shop. This is done through data governance, both through hierarchies and attribution. Consider online shopping behaviors: is a shopper looking for women’s apparel? If so, is it tops or bottoms? If bottoms, is it a pant or a skirt? The data in the background needs to be available to support the breadcrumbs of getting your shopper to their options quickly and without frustration.

Further, there is a balance between having strong data governance and being too strict. As an example, many products are grouped for pricing reasons. Most retailers will price all Coca-Cola products consistently, regardless of the demand for each flavor. In a recent client price optimization implementation project, those groupings didn’t exist in master data because when the retailer replaced the core systems, they didn’t realize the need. The groupings were created and maintained in Excel, but because of an overly strict governance process, it would take months to rectify the situation so the appropriate price groupings could be used. Net-net, you can take the data governance concept too far, and so companies need a blend of structure plus agility when managing master data.