The rise of automated commerce is already reshaping the customer experience. But is it the future of retail?
BY JESSE DONALDSON
FOR forward-thinking retailers, a-commerce is the new e-commerce. Born out of the intersection of big data, AI, the Internet of Things, and predictive analytics, automated commerce (not to be confused with augmented commerce) is a set of innovations that could transform the customer experience in the decades to come, not only allowing consumers to outsource key purchasing decisions, but also helping retailers to improve efficiency, products, and services—not just instore, but behind-the-scenes, online, and even right in the home.
store, but behind-the-scenes, online, and even right in the home. “There are a lot of applications,” notes Daniel Moneta, EVP of Corporate Development at MMB Networks, an IoT Connectivity company. “There are warehouses with robots picking up things and bringing them to packers. We’re seeing retail locations they’re experimenting with, where they don’t even have cash registers. You just walk in, scan with your phone, grab what you want, and leave.”
“WHAT WE’RE WORKING ON IS TAKING ADDITIONAL SETS OF DATA; UNSTRUCTURED DATA—THE VOICE OF THE CLIENT, WEATHER DATA, IOT SENSOR DATA FROM A REAL-TIME PERSPECTIVE. WE’RE USING IT TOGETHER, AND THEN THESE MODELS BEGIN TO LEARN, AND TO OPTIMIZE.”CHARBEL SAFADI
“Retailers are increasingly adopting technology that has advanced retail science baked in,” adds Jeff Warren, VP of Strategy at solutions firm Oracle Retail, “and that can improve the overall customer experience by better anticipating customers’ next purchases. In practical terms, retailers are generating more accurate forecasts, reducing inventory levels and reducing markdowns by leveraging planning and demand forecasting solutions with embedded AI. Also, predictive analytics and AI can provide more personalized and relevant communications to consumers—supporting a faster and more tailored shopping experience. Finally, connected devices can streamline the customer experience outside the store, with IoT-enabled reordering functions or via voice assistants capable of quickly placing orders or carrying out changes to orders.”
Descriptive, predictive, and prescriptive
While a-commerce is still relatively new, it is simply the next step in a digital revolution that has both disrupted and reshaped the retail ecosystem over the past two decades, building on developments in online shopping and data collection. And online is where a-commerce’s potential is perhaps the most obvious, allowing today’s retailers to automate virtually every stage of the purchase journey—from order management, to order fulfillment, to recommendations. And now, thanks to advances in AI and machine learning, online shoppers can even take part in a digital version of the in-store experience; known as conversational commerce, AI can now provide consumers with a digital assistant capable of understanding their purchase history and their needs, and field their questions in real-time.
“We expect that conversational commerce will be widely embraced across critical activities such as order management—specifically with order inquiry and maintenance,” Warren says. “Automation in back-end operations—through direct integration of digital assistants with core enterprise applications—allows retailers to automate mundane tasks. Case in point: with direct integration between planning systems like offer optimization to digital assistants, retailers can automate promotions, targeted offers and markdowns to engage omnichannel customers with personalized offers. Additionally, with respect to customer service, digital assistants provide the first line of defence for questions and mundane tasks that service representatives would otherwise execute countless times per day.”
However, one of the biggest a-commerce innovations is one that, from a customer perspective, has taken place mostly behind the scenes: automated demand forecasting. This technology is capable of using AI and machine learning to predict customer demand before it even happens. And companies that have adopted it are already reaping huge benefits. According to a 2017 survey by the McKinsey Global Institute, proactive AI adopters in the retail sphere have already seen their profits rise 15 per cent above the industry average. The study predicts that in the years to come, demand forecasting alone could reduce transportation costs by 5-10 per cent, warehousing costs by 25 per cent, and lead to a massive 65 per cent reduction in sales lost due to unavailable product. German-based retailer Otto is using demand forecasting to determine their whole inventory, something that has already been accurate up to 90 per cent of the time. Automated demand forecasting is different from traditional models in that it allows machines to automatically interpret all manner of data—not just customer purchasing habits, but a complex ecosystem of in-store, online, and external factors.
Proactive AI adopters in the retail sphere have experienced a profit rise 15 percent above the industry average.
Demand forecasting could reduce transportation costs by 5-10 percent, warehousing costs by 25 percent, and lost sales by 65 percent.
21%Source: McKinsey Global Institute 2017 survey
Percentage of North American shoppers who are willing to allow retailers access to their mobile devices in exchange for tailored communication.
“AI and machine learning have changed the approach fundamentally,” explains Charbel Safadi, Canadian AI Team Leader at IBM. “Historical data is important, but that is only one set of data points that organizations have been using to predict the future. What we’re working on is taking additional sets of data; unstructured data—the voice of the client, weather data, IOT sensor data from a realtime perspective. We’re using it together, and then these models begin to learn, and to optimize.”
“Advanced analytical tools allow retailers not only to understand what happened, but also look forward to what might happen, and what action to take to maximize business value,” Warren says. “It can be descriptive, predictive, and prescriptive. IoT and big data play an essential role in both understanding demand and automating fulfillment.”
And, as Oracle’s research demonstrates, an increasing number of consumers are not only embracing demand forecasting as a concept, but many seem willing to share their data in exchange for a more personalized retail experience.
“We commissioned a research study this fall that examined emerging trends across 15 countries,” explains Warren. “We asked consumers what their appetite was for push communications from their grocery and pharmacy retailers based on personal device data. We found that the interest is there— albeit more prevalent in Latin America and the Middle East versus North America or Europe.”
According to Oracle’s research, 21 per cent of North American and 26 per cent of European shoppers would be willing to allow retailers access to their device data in exchange for communication tailored to their needs. But in Latin America and the Middle East, that number rose to 49 per cent and 56 per cent, respectively. However, as a study conducted by analytics and tech company Reply cautions, that reach only goes so far; while today’s consumers are eager to integrate technology into their shopping experience, they’re only comfortable with retailers automating the purchasing process, not the purchasing decision itself.
“Automation needs to be understood in a way that considers users’ feelings of individuality and uniqueness,” the report warns. “No one wants to believe that a machine can accurately predict their behaviour or the decisions they make. In general, users showed strong reluctance to entrust a machine with decisions they consider to be genuinely their own while they value the feeling to stay in control.”
Minding the store
Luckily, when it comes to automating the purchasing process, many retailers are on board— something that is even happening in-store. One of the most talked-about examples of brick-and-mortar a-commerce is Amazon Go. Back in January of 2018, Amazon opened its first cashierless retail location, an 1,800-square-foot facility in Seattle that uses AI and computer vision to log customer purchases, and charge credit cards without any need for a physical checkout. And while the technology itself has yet to be adopted at scale, it’s something consumers are excited about; according to a recent SOTI study, 73 percent of consumers want mPos for quicker checkout times, and 61 percent would use a self-serve kiosk instead of speaking to a sales associate. But in the world beyond Amazon, a-commerce has slowly been integrating with brick-and-mortar locations in all manner of other ways.
In 2013, Starbucks began connecting its fridges and Clover coffee machines to the internet, allowing them to digitally update recipes and information in real-time. In 2017, UK grocery retailer Tesco trialed an automated delivery service that was able to deliver groceries via a six-wheeled robot to anywhere within a three-mile radius within an hour. Back in 2015, Target was one of the first retailers to introduce a digital shopping assistant—a smartphone app that uses GPS technology to aid in the purchase journey, giving customers directions, bringing up comparison information for products they approach, and automatically alerting consumers to promotions nearby. And as Safadi notes, companies like IBM and Cortexica are taking the Shopping Assistant concept even further, integrating computer vision with mobile technology to provide automatic recommendations to consumers, based on nothing more than image data.
“A consumer can take a picture or upload a snapshot of a person and we’re able to understand and classify any object in that image,” he says. “We can tell where the shoes are and even extract the image of that shoe. Then we use vision technology—which is entirely AI-driven—to understand the characteristics of that shoe and use that metadata to match it with our available catalogue, dynamically bridging the gap between the consumer’s behaviour and our product inventory.”
In the future, these innovations may also be common in our homes. The smart appliance market—worth an estimated $20 billion—is developing thermostats and appliances capable of learning consumer habits and tailoring their usage accordingly. Smart fridges will be able to detect when food has gone bad, and place online orders for delivery. Much of this is still years away —and many of the acommerce innovations at the retail level have yet to be adopted at scale. But, Warren says, as the years pass, as costs come down and infrastructure grows, and as more and more retailers—be they small, large, or mid-size—join the a-commerce revolution, adoption may end up being all but automatic.
“Increased efficiency, intelligent starting points, reduction in mundane tasks—these are all advantages of automation,” he says. “We also see more consistency between strategy and execution, simplified decision-making and maximized accuracy and scale using artificial intelligence, machine learning, and decision sciences. The net effect is that retailers can free up their personnel and funnel those resources toward managing more complex projects and more creative endeavours.”