The American Marketing Association has defined marketing as “the activity, set of institutions, and processes for creating, communicating, delivering, and exchanging offerings that have value for customers, clients, partners, and society at large.” It is clear by its definition that marketing lies at the heart of the success of a business and is an age old mechanism to make sure that a company is selling the right product, at the right price, running the right promotions and selling at the right place via the most optimal distribution channels. Product, Price, Promotion, and Place are therefore rightly called the 4 P’s of marketing. Traditionally, decisions around these 4 P’s were based on experience and the “gut-feel” of marketing managers. While no one can discount these important factors, the vast amount of data that firms have at their disposal today has started to play an ever increasing role in helping managers make marketing decisions. This is especially true in the retail industry which has the privilege of direct contact with their customers and hence, compared to many other industries, has access to rich customer data. This paper examines how marketers, especially in the “Big Box” retail industry, are leveraging the data for better business outcomes.
Let us start with the first ‘P’ of Marketing – the Product. For Big Box retailers, determining the products that it should keep on its shelves, ability to forecast product sales, and bringing efficiency to the product supply chain are important aspects of marketing. Retailers have started to use data in each of these aspects to drive growth and profitability. For example, data collected from RFID tags, geofenced stores, headcount cameras, and gesture recognizing movements help retailers understand the pattern around the paths customers take within their stores, determine which products get picked up by customers but are put back on shelves, determine rates at which products disappear from shelves, and decipher patterns around product returns. This data is then used in conjunction with that collected from sales channels such as online, social, and mobile to optimize the product mix the retailer should keep on its shelves. Big Box retailers such as Walmart and Meijer also use this data for optimal placement of products in their stores and on shelves.
Making sales forecasts is another area where data plays a crucial role in helping retail marketers make product decisions. Sales forecasts are important for a variety of reasons including the ability to keep a Just-In-Time inventory, avoiding out-of-stock situations, and situations where inventory keeps sitting on shelves. Sales forecasts were traditionally done based on gut-feel of marketing managers and store clerks. They now augment the gut-feel with hard facts and figures emanating from data collected across the various sales channels. Sales forecast techniques such as Time Series Forecast models, and Smoothing models are fueled by internal and purchased data.
Sales and revenue forecasts made using the analytical methods described above, however, are not able to explain variances from the actual observations. Both internal and external factors contribute to these variances. Retail marketers are often tasked with identifying these factors, determining their contribution towards the variances observed, and ways to incorporate the factors in models going forward. The technique of Business Driver Decomposition (BDD) comes handy here. BDD is based on marketing mix models that allow for controllable internal factors such as marketing, distribution, price, promotion, etc. to be modeled together with external factors that the company cannot control such as seasonality, macro-economic conditions, competitors etc. The approach usually enables marketers to run simulations and perform what-if scenarios under different combinations of internal and external factors. These simulations assist in determining the most optimal mix of products to sell.
The second ‘P’ of marketing is Price. It is absolutely important for marketers in various industries including retail to price their products optimally. A higher than the “right” price can drive customers away. Conversely, a price point below the “right” price can erode profitability. What makes pricing much more complex in the retail sector, is that it needs to be optimized across sales channels, stores, and customer segments.
Various analytical techniques, which largely depend on data, are leveraged by retailers today to derive insights by taking the traditional methods of customer segmentation forward and performing micro-segmentation. The holy grail of micro-segmentation is to be able to segment the market at the individual level and tailor offerings & pricing for individuals rather than for broader segments. While pricing at the individual level is still some way out (and its opaqueness may even cause backlash from customers), marketers in retail use sales data to derive valuable insights by clustering stores into zones, assessing shopping behavior across channels, and measuring differences in demand across customer segments. These insights allow retailers to define a framework to set prices while accounting for differences across customers, channels, competitors and product categories. The source of most of the sales data is online transactions on the company’s e-commerce website(s), PoS systems collected in stores, loyalty programs, and social media activity.
The practice of dynamic pricing is on the rise all around us. Companies in the airline industry were pioneers in the area but other industries including retail have caught up. Dynamic pricing allows retailers to adjust pricing in near real time by taking into account a variety of data points emanating from sales, supply and demand analysis, competitive intelligence, market conditions, weather conditions, and local sporting events. Sears, for example, changed prices on 20% of their products at least once a day during the 2014 holiday season.
The third ‘P’ of marketing is Promotion. Promotion means the various ways in which marketers present their product and its differentiating characteristics to customers. This is usually done via methods such as advertising, and public relations using a variety of media including print, digital, and social.
Promotion in the Big Box retail industry is particularly hard since retailers do not produce the products they sell and therefore it is not easy to provide differentiation. They have to use other ways to differentiate from competition. One of the ways is the ability to reach the right customer, at the right time, with the right offer, and on the right channel. Knowing one’s customers better is a prerequisite to build this ability. The best way to know the customers is to leverage data they generate both offline via loyalty programs, call center interactions & PoS systems and online via social media and company website(s). The data includes customers’ clickstream data, purchase history, product reviews, social media activity, demographics, and financial situation. Once marketers know their customers, they understand their particular likes and dislikes, needs and wants. The information, in turn, is used to generate the “Next Best Offer (NBO)” to individual customers. NBOs are very powerful since they are contextual, tailored and relevant to a particular customer. Tesco, a UK based retailer, uses NBO as a strategy to increase sales to its regular customers with targeted coupon delivered through its loyalty program. For example, Tesco’s loyalty program members that buy baby diapers are sent coupons for not only baby wipes but also for beer as data analytics has revealed that new fathers spend less time at the pub and hence tend to buy more beer.
Retail marketers take the NBO to their next level of maturity when the location information of the customer is combined with the data described above to make the offer even more relevant and immediately executable. For example, Target, Walgreens and Walmart use Bluetooth Low Energy (BLE) beacons to identify the location of a customer in a store, combine the information with the purchase history of the customer on various channels to determine the next best offer and push relevant specials and discount coupons on their mobile devices.
Marketers in the Big Box retail also use data for Customer Engagement Assessment to determine underserved customer groups with the likelihood of spending more. Usually data from PoS systems, and online activity combined with demographics of customers is used to help with such an assessment. Targeting and promoting to the underserved groups is generally a profitable strategy that retailers follow. Data from identified sources along with data purchased from external sources is usually leveraged to determine Share of Wallet that a particular firm enjoys from a particular customer segment. This data can be useful in finding out where the share is being lost, to whom and why.
The fourth ‘P’ of marketing is Place. Place means the distribution strategy that a particular company employs to reach the customer at the right place. Marketers in the retail industry use sales and inventory data from multiple source systems to identify items at risk of stocking out or of being overstocked. This is important especially in modern day retail business where customers are given the option to order online and pickup from physical stores. Providing that seamless experience to customers is important to compete effectively in the marketplace. Also, predictive and prescriptive analytics are leveraged to not only predict demand at various consumption points but also to optimize the allocation of inventory optimally across these consumption points.
In this paper, we saw how marketers in the Big Box retail sector use data collected from a variety of sources to make marketing decisions which until a few years ago were made based purely on intuition, and experience. In the modern retail marketplace, however, where customer has access to information on her fingertips, has all sorts of options to buy, and has the power to sully a brand without a lot of effort, relying just on intuition and experience to make decisions around product, price, promotion, and place can prove fatal to a retail organization. On the other hand, using data to make the decisions can be the differentiator that retailers constantly seek. But one must understand that it takes commitment, and persistence at all levels within an organization – most importantly from the CEO – to make the organization a data driven organization.
About Ashish Kashyap
Ashish Kashyap is a Senior IT Leader with over 15 years of software consulting experience in a variety of industries such as pharmaceutical, retail, airline, automobile, marketing, and utility. He is an experienced solutions architect and has designed & developed B2B and B2C kind of applications on a variety of platforms including J2EE, COTS, and SAAS.
More on Ashish Kashyap.