Retail is only scratching the surface of the potential of analytics
By Shawn Smith, Retail Solutions Specialist, SAS Canada
The customer is always right.—Gordon Selfridge, founder of Selfridge’s department store, 1909. Those iconic words have guided the retail industry for more than a century (though there are some that argue they have misguided the retail industry). Regardless of which side of that particular coin you choose, it’s no secret that customers want better service, and they’ll repay it with their hard=earned dollars.
What’s different about today’s shopper is that he or she is more empowered than ever. Social media can affirm or condemn retailers’ reputations. Product reviews are instantaneously available. Comparison shopping sites deliver the best price from dozens of retailers. And the technology that may be transforming the customer experience most—mobile devices—puts all of that information into their hands at the point of sale.
What’s different about today’s shopper is that he or she is more empowered than ever. Social media can affirm or condemn retailers’ reputations.
Senior Solutions Specialist, SAS Canada Retail Practice
The retail customer has many roles today, as opposed to someone who simply puts cash in your till. They are collaborators who use social networks to advise and be advised. They are voracious content consumers. They are discerning price and service comparators. And they’re demanding around-the-clock service through multiple channels.
The changing consumer is the challenge that retail faces. Fortunately, though, we have a tool to help meet that challenge in the massive volumes of data that are created and autonomously collected by these transactions, whether they’re point-of-sale or on social media. And we have technology in the form of huge processing power and analytical software to turn that data into actionable information that will allow the retailer to deliver the right offer to the right consumer at the right time.
A HISTORY OF ANALYTICS
The retail vertical has been among those at the forefront of analytics since before it was called analytics. Merchandising decisions—what makes for good impulse buy item, how do we position items for maximum exposure, what products do we group together? What do we offer on sale, and at what point in its life cycle, or at what seasonal inflection? How do I manage the supply chain and inventory issues? A lot of data was collected, but the analytical engine was management’s gut feel based on those numbers. Great retailers had a culture of great instinct. Instinct is still a powerful strategic differentiator, possibly still the most important. But it’s now supported by data capture and analytics technology that provides vastly richer and more easily consumed information to make those strategic decisions.
For the data to provide that strategic support, a certain amount of retooling is required. Driving value from data is built on five pillars.
*Data. Obviously, an evidence-based decision-making process is built on data. It’s actually easier for retail than most verticals because of the high volume of structured transactional data—clean, rich, diverse, culled from point-of-sale transactions and loyalty programs and myriad other easily defined data sources. It’s actually harder for retail because of the volume of unstructured data—no other vertical attracts more social chatter.
* People. The most important pillar, where instinct meets technology. A business needs people with business skillsets, who can visualize the relationships among various data. Right now, there’s a huge demand for post-graduate data scientists. They’re rare, they’re expensive talent, and the competition for them is a new frontier for recruiting. That kind of talent should be reserved for strategic tasks. But those with hands-dirty business acumen should also have the tools to make tactical, operational, and logistical decisions based on enterprise data. We call these roles the citizen data scientist; those close enough to the ground to know what questions to ask, to determine relationships among datasets. They’re an invaluable resource—if they have accessible tools to glean insight from them.
* Technology. There are two tools critical to enabling the citizen data scientist. One is the hardware and software power to allow the processing of vast volumes of data in near-real-time; many of these are based on open source technologies like Hadoop and MapReduce. (You can learn more about these in this article on our web site.) The other is analytics tools that are usable by business talent: visual graphical user interfaces (GUIs) that are intuitive to someone who is not a technologist, forecasting tools, predictive analytics. That’s the foundation on which a culture of business analytics lies.
* Culture. If business users don’t work with the data collected, if they don’t see the value and apply their specialized skill to drive insight, then analytics is a waste of time, money and data. Retailers must create a culture in which using data in an analytical context is exciting and rewarding. Once citizen data scientists understand the power they have in their hands to drive business outcomes—revenue, supply chain efficiency, more effective marketing and merchandising—they’ll become more engaged in their roles because they will be producing results.
*Process. Workflow isn’t exciting for everyone, but it’s a foundation that allows the other four pillars to produce results. Data must be collected cleanly and be accessible for users. Data scientists must determine how unstructured data is dealt with. Citizen data scientists must have clear protocols so they can unleash their subject matter expertise.
WHAT CAN WE DO WITH THIS DATA?
Analytics can help support those gut-level decisions and drive the process into the 21st Century. For example:
* Affinity. It’s always been a merchandising and marketing preoccupation. What products go together, whether on a shop floor or in an online shopping basket? Many instinctive connections can are confirmed by a recent MBA study of affinity in grocery stores: bacon and eggs; pancakes and syrup; pasta sauce and, well, pasta. But would you have thought of dog food and household cleaning products (dogs are messy, after all)? Peanut butter and yoghurt? Prepared salads and rotisserie style chicken? Precooked dinners and granola? There are strong affinity patterns between these purchases that can help guide decisions about placement in a store or cross-marketing promotions (get 50 per cent off carpet cleaner with your purchase of kibble—please!).
(By the way, we now have scientific proof that things do go together like peanut butter and jam, to the extent of a 5.526% lift.)
* Marketing. In conjunction with loyalty program data, analytics can create a profile of purchasing patterns of repeat customers. This, combined with mobile technology like beaconing and smart phones, can lead to the Holy Grail of marketing: The ability to give a customer the right offer, at the right time, in the right place. In the future, sophisticated retailers will be able to change the shelf price of items dynamically.
A caveat: Protection of customers’ personal information will be a critical concern, as the Liquor Control Board of Ontario (LCBO) recently found when a court ruled it was collecting data that infringed on the privacy rights of members of its wine club. Any organization that uses customer information to provide customer offers need a vigilant corporate security officer or corporate privacy officer to ensure it’s not overreaching on the personal data collection side.
* Supply chain management. Just like it can help deliver marketing offers, data can help make sure the right product is in the right store at the right time. SCM is an established discipline, being a long-time partner in the enterprise resource management scheme.
* Loss prevention. A huge bottom-line factor for retailers, loss prevention—from third-party theft, fraud and shrinkage—can make a large difference. Analytics can help loss prevention managers isolate the sources of such abnormalities and design changes in process to reduce losses.
SCRATCHING THE SURFACE Retailers are among the leaders in the application of the science of data use to business problems, and have been for many years. But we’re only scratching the surface of the possibilities that analytics offers. From the shop floor to the back office to the warehouse, here are analytical applications that can improve your company’s competitive position in a retail world that’s becoming more competitive every day.
Originally published in Canadian Retailer Magazine Holiday 2015 issue
Senior Solutions Specialist, SAS Canada Retail Practice
Shawn Smith is a Senior Solutions Specialist in SAS Canada’s Retail Practice. In his current role, he is responsible for helping clients gain value from current and future investments in analytic retail technologies. Shawn has extensive knowledge and expertise in retail analytics, and in devising and executing strategies to better understand shopper behaviour and omni-channel evolution.