In recent months, big data has been an often-discussed topic among business leaders as they consider ways to adjust their businesses to contend with the effects of the pandemic, which continues to shape shopping behavior. Across the retail industry, companies are still struggling to put data-driven strategies into action.
In November 2020, Exasol published research titled “Retail: Decision making during times of uncertainty,” which revealed that since the pandemic started, 87% of retail organizations in the U.S. have been under pressure to make data-driven decisions faster. The research showed that 82% of retail organizations in the U.S. are data-driven businesses, yet more than half (61%) of organizations say their current data infrastructure does not allow them to innovate swiftly enough. In fact, more than half (52%) of retail organizations reported their need to invest in data infrastructure for faster insights.
In today’s environment it has become vitally important to become data-driven and intentional about each and every business decision. And among the many benefits derived by implementing data-centered strategies, is that organizations that embrace a data-first mindset can apply the resulting insights and develop business actions and strategies more confidently and quickly.
According to Julie Bonnell, vice president of operations at HRG, five big changes in the area of data application and its management have emerged in the past 20 years.
Omnichannel — online presence has become a necessity, with many purchase decisions being made in advance of a shopping trip; expectations for congruence between brick-and-mortar and the product data and images on e-commerce sites is imperative.
Micro-fulfillment — at the center of this phenomenon — the idea of placing small-scale warehouse facilities in accessible urban locations or inside a reconfigured retail operation close to the end consumer who is making a purchase — is data. From assortment planning and placement to analysis of inventory turns and improved shopper efficiency (the number of seconds per one dollar spent on the shopping trip), retailers without a well-thought-out data management infrastructure or plan are likely to underperform.
Privacy — heightened protection and anonymization of data has become a requirement as the risk of breaches has risen sharply.
Personalization/hyper-localization — decisions about the retail assortment must align with customer need, and a delicate balance is required between shopper satisfaction and profit. Point-of-sale data analysis will reveal how to reach that balance.
Attribution — classifying, categorizing and stratifying data leads to analysis with a more granular vantage point to help identify trends and more easily convert them to growth opportunities. This drives greater assortment rationalization and shopper personalization at retail.
The benefits of a data-driven retail strategy include such things as cost efficiencies and reduced waste, and speed and agility in modifying product mix, leading to customer experience improvements, easier loyalty program management and increased personalization, to name a few. Decisions must first be made regarding the mountains of potential data inputs and their prioritization, but once determined, the results can be incredibly impactful.
By analyzing trends, purchase history, past behavior and other critical data points, retailers can quite accurately identify the key factors that contribute to a purchase decision. Equipped with these insights, future customer behavior can be more accurately predicted, and adjustments made to overall marketing efforts, loyalty program elements, service offerings and promotional efforts.
Retailers who actively monitor and analyze consumer trends can stay ahead of their competition to maintain their status as the “go-to” location for the latest or most popular items. Through analyzing buying behaviors and testing new methods, retailers can spot opportunities to build consumer loyalty or capitalize on the growing popularity of a product or category (e.g., a recent example is the number of retailers who are grouping personal safety, disinfection, and health and wellness items into a designated category. Global Market Development Center (GMDC) recently published a white paper on this subject: https://gmdc.org/personal-safety-essentials-ppe-disinfectants-hbw/).
Another application of effective data mining is in the area of pricing strategy. It is no longer sufficient to set pricing based on revenue goals or gut feel — data must be at the center of these decisions. Since the mid-’80s, HRG has focused on getting the “base pricing” right before applying price elasticity software, machine learning approaches or promotional pricing science. All price strategies must begin with a solid and reliable base pricing methodology.
In-store product displays can be better optimized when data drives decisions. Optimal in-store product placement, as well as category adjacencies, can be enhanced through data analysis. As an example, grouping curated product assortments into an easily accessible category requires careful examination of market basket and new, cross-category logic.
Data and analytics should be used to enhance visibility into a retailer’s supply chain, inventory and operations, as well as transform short-term forecasting and modeling. The pace of today’s market demands quick decisions fed by accurate data. As an example, in the area of markdowns, retailers are using data to compress performance thresholds for promotions and new-to-market brands.
Technology should be viewed as an enabler. Turning once again to HRG’s Bonnell, her advice falls into three key areas: data harmonization, data fitness and data relevance.
Data harmonization means aligning data sources for consistency and analysis. Harmonizing does not mean homogenizing, rather viewing the data through one lens. Bonnell suggests that retailers begin by answering some tough questions.
How many sources of data are required, and why are they being selected?
Are you using the right variety of sources?
What will be the source of truth?
Can you effectively eliminate self-fulfilling prophecies and data biases?
Data-driven organizations must determine their ultimate goal regarding data. Presuming that the direction is often a function of cost — few, if any retailers, have dollars to build an ideal state — they must make do with the data that is available. Sometimes striving for excellence rather than perfection allows you to move more quickly. As Bonnell suggests, “You have to decide whether you are creating a solution that is fit for purpose or best in breed.”
Developing plans to maintain data currency and manage expectations begins with a definition of “current.” As an example, when planning their assortment, a retailer’s goal may be to limit the purview to include only items currently available in the marketplace. The challenge, however, is considering what to do with inventory still on shelves or in a warehouse that is potentially supported by obsolete data. Beginning with the end in mind and remaining proactive with data choices and definitions is vital.
Committing to using data for the basis of decisions is not enough. Facts and figures are meaningless if an organization cannot gain valuable insights that lead to more-informed actions. The pandemic has revealed an exasperating problem for most retailers: an information deficit and the inability to quickly and effectively predict new trends from the most recent data. As a result, retailers must look beyond traditional point-of-sale transaction or shopper loyalty data. As Albert Einstein suggested, “Make everything as simple as possible, but not simpler.”
Dave Wendland is HRG vice president of strategic relations. HRG focuses on improving results across the retail supply chain by addressing dynamic needs such as assortment planning and placement, retail execution strategy, fixture coordination, item database management, brand marketing, Rx track-and-trace compliance, and analytics. He can be contacted at email@example.com.