Retail AI: adopt or die moment

Companies can increase their profit in two ways: reduce costs or sell more.

According to Business Insider, several well-known retailers, including Payless ShoeSource, The Limited and RadioShack, have recently gone bankrupt [1]. This downward spiral isn’t likely to stop anytime soon. At a time when the population is increasing, and customers’ needs and wants are rapidly growing, why is the retail industry shrinking? There are no shortages of reasons (see inset). 

Smart Machines to the Rescue

Fortunately, help is on the way thanks to exponential progress in “cognitive computing” – a close cousin of artificial intelligence, machine learning and other self-learning systems fueled by the combination of cloud, big data, mobile, Internet of Everything and new algorithms that use x-mining, pattern recognition and natural language processing to mimic the way the human brain works. 

Smart machines are becoming more like humans through their ability to recognize voices, process natural language and learn. They learn by accepting feedback and interacting with the physical world through devices that enable them to see, hear, smell and touch, as well as through mobility and motor control. They understand structured and unstructured data, and they do a much faster and better job than humans at recognizing patterns, performing rule-based analysis on very large amounts of data, and solving both structured and unstructured problems. 

Smart machines are rapidly transforming the fundamental nature of a wide range of public and private organizations and revitalizing these organizations’ digital business models in retail and consumer packaged goods. This revolution provides the means to improve the efficiency, effectiveness, sustainability and innovativeness of product and service offerings. 

Let’s discuss some of these new ways of utilizing smart machines.

Demand forecasting. One of the major problems in the retail industry is accurately calculating supply and demand. By analyzing large amount of data in much shorter duration, cognitive tools can predict future supply-demand scenarios much more accurately than any human. For example, Obase/DTLR’s (https://www.dtlr.us/) demand forecasting model embodies machine learning algorithms that can compute the demand for a product in a store for that day, week or month. 

New supplier identification. For those new to sourcing in a category, a cognitive tool can pull information into an easily digestible format to help identify potential new suppliers, run objective request for information assessments and identify risk factors. The system can analyze supplier consolidation opportunities, savings/payment term opportunities and costs of noncompliance, as well as assess reliability and performance issues on products and services. 

Production planning. Smart machines can adjust production planning by predicting, for instance, maintenance needs and anticipating bottlenecks relative to inventory/delivery. With risk management, sometimes the information that would help avoid some type of loss is not information that users knew they should be paying attention to. Sensors are replacing human hands, resulting in less wasted time and materials as well as optimal accuracy and workflow, lower production costs, faster turnarounds and more efficiently meeting customer demand [2]. 

Contract management. These systems could become a robust repository for contracts and agreements, analyzing contracts for performance versus terms, and instigating corrective action, renegotiation or other supplier management activities. Imagine an organization that has acquired another company and all of its contracts. Cognitive tools can analyze those activities to capture the firm’s standards and drive contract personnel activity to address areas of concern. 

Marketing. Using smart machines, campaign management can be integrated with physical stores POS, kiosk, e-commerce and mobile commerce applications, providing an omni-channel experience that can be easily created by business users via software wizards. Through the use of award types such as discounts, coupons, checks, points or messages, DTLR’s (footwear, apparel and accessories retailer) campaign management program provides holistic, targeted campaigns with customer loyalty programs. 

In-store traffic management. Gesture recognition solutions can track specific visitor activity, but less intrusive means of tracking have also been developed. For example, floor-level cameras can track traffic and where people spend time in stores, and an analysis of the video can predict the age and gender of customers interested in certain shoes. IBM’s work with cognitive computing has produced things such as natural language analytics that can judge vocalized consumer reaction to products, as well as humanoid robots that can chat with customers. 

Customer experience. Cognitive computing models can use data sets to predict and prioritize the most successful campaigns and channels and provide these insights to decision-makers. Using Google’s Tango 3D enhanced reality app, Walgreens is testing an application that shows customers where the products they are looking for are located in a store. Amazon Echo can order from home via the Alexa sound front. 

Workforce optimization. Workforce is one of the most expensive components of retail operations. DTLR’s workforce management is a data-driven solution, forecasting the number of full- and part-time staff required to satisfy target customer service levels and to get the work done by utilizing algorithms that consider all the correlated factors to generate the forecasts: sales, products, team, activities, assignments, workload rules, etc. 

Logistics and delivery. Domino’s Robotic Unit (DRU) can keep food and drinks at the appropriate temperature; the DRU’s sensors help it navigate a best-travel path for delivery. Amazon drones might become a future means of safely delivering up to five-pound packages in less than 30 minutes.

In-store help. Watch out – a smart machine might steal your job!

Other retailers are using big data to reimagine how their stores help shoppers find what they need. Target, which unveiled a multibillion-dollar plan early this year to play catch-up in e-commerce, is planning to arm floor associates with new technology to help visitors. The new point-of-sale system will be able to search real-time inventory across the organization, arrange for shipping and take payment from the customer. 

Another example is Amazon Go, whose slogan is, “No queue, no checkout.” The “just walk out technology” required to make checkouts and cashiers obsolete involves a mix of sensors and artificial intelligence to determine what products shoppers pick up off the shelves.

Home improvement chain Lowe’s has taken things one step further with mobile LoweBots. They are able to answer simple questions from customers and locate items as they roam around the store. They also keep track of inventory and track shopper patterns to help the company make better business decisions. 

Other common application areas that are becoming popular include a solution for information exchange using an automated human agent (i.e., chatbot) that has the ability to develop a conversation using natural languages or to provide a virtual tour of the physical store. Macy’s is collaborating with IBM Watson to pilot “Macy’s On Call,” an intelligent chatbot-based program that will answer customer questions via their mobile devices as they shop in-store.

These are just a few examples of how smart machines are reshaping the in-store customer experience in the retail industry; many more are on the way. 

What’s Next?

We believe a comprehensive reimagining of business strategy and operational capabilities is necessary for retailers to chart a course for success in the future. However, new strategies, technological innovations and customer-centric omni-channel operating models should be combined with old-fashioned business acumen.

While the promise of smart machines is to change the retail and omni-channel operation, business leaders are a bit slow to adopt this into their day-to-day businesses. Even in a fast-paced industry like retail, there are more proof-of-concepts than end-to-end business solutions offering cognitive capabilities. Brick-and-mortar stores need to be more engaging, more personable and more human than any online retail experience can be, with or without cognitive helpers. 

Given the hype and opportunities, many major technology platform vendors and start-up companies are jumping on the AI/cognitive computing bandwagon. So, where do we start?

First, the goal of any retailer should be to transform and accelerate its business using cognitive computing to optimize business processes, speed up processes and eliminate friction for core business functions. This usually focuses on reducing costs and increasing sales. What should a cognitive computing roadmap look like to achieve such business objectives? 

Second, it starts with data [3]. Data is digital gold, and data inequality will prove a major battleground, but most businesses have dark data that offer little insight. Better customer experience and optimization across all business functions will require an ever-growing pipeline of data collection that will demand new hardware, software and networking investments.

Third, think about insight generation. Using cognitive computing techniques to process large volumes of data to develop actionable insights that would not have been possible with previous approaches. This is about scale, giving businesses the ability to generate insights at a scale that was not previously possible.

Fourth, customer engagement; leverage the power of cognitive computing to engage with customers in new and more effective ways. This enables entirely new customer experiences and touch-points and dramatically improve existing ones [4]. 

Does the retailer have the appropriate skills and resources in-house to embark on this journey? The mindset will expand from a “What do I sell?” perspective, where the primary focus is on physical products, to a “How can I better serve customers?” perspective about value co-creation and service. In looking at the future, a retailer needs to consider these questions:

Do you know who your customers are? Can you change the way the business operates to better serve the customers? Do you really know what your customers want? Are you able to shape their expectations? Can you recognize disruptive threats, as well as opportunities, before they happen? What safeguards have you built for your business? Are you capable of adapting your business model and collaborating with customers and colleagues in a new retail ecosystem?

While it’s impossible to know precisely what retail will look like moving ahead, we believe there will be a paradigm shift, changing the nature of retail. The overall goal of cognitive computing is to increase the productivity and creativity (decision-making, connectivity, innovation and augmentation) of individuals and organizations [5]. 

This is not about digital transformation. It is about retail service transformation partially enabled by digital and smart machines. Today, when a customer buys a drill, does he or she want a drill or a hole? According to research, people, don’t want to buy a quarter-inch drill. They want a quarter-inch hole. Robert Lutz, chairman of GM, put it this way: “An automobile is actually art, entertainment and mobile sculpture, which, coincidently, also happens to provide transportation.” 

Today:

• Customers want to “hire” a product to do a job.
• Commoditization of products results in price and margin pressures.
• Customers are demanding services and solutions.
• Services can provide platforms for profitability.
• Loyalty and customer satisfaction are often driven by services.
• Service offerings can differentiate firms in highly competitive industries. 

Acknowledgment

Part of this article is excerpted with permission of the Publisher, HBR Turkey from Demirkan, H. and Dal, B., “Smart Machines and Consumer Goods,” Harvard Business Review, Turkish Edition (published in Turkish), February 2018.

 References

1. https://newjerseybankruptcynow.com/retail-industry-decline/
2. Newman, D., 2017, “Top 5 Digital Transformation Trends in Manufacturing,” Forbes, Aug. 8 (https://www.forbes.com/sites/danielnewman/2017/08/08/top-5-digital-transformation-trends-in-manufacturing/#4b50b2d9249f).
3. Demirkan, H. and Dal, B., 2014, “Why Do So Many Analytics Projects Still Fail? Key considerations for deep analytics on big data, learning and insights,” Analytics, July-August issue, pp. 44-52.
4. Demirkan, H. and Dal, B., 2016, “Get Smart: Digital business innovation,” Analytics, January/February issue, pp. 40-48.
5. H. Demirkan, J.C. Spohrer, and J.J. Welser, 2016, “Digital Innovation and Strategic Transformation,” IT Professional, Vol. 18, No. 6, pp. 14-18.