March of the Retail Robots
Smart machines will upgrade the in-store experience, as well as the business model
by: Martin Vilaboy
On initial walk up, SCOTTeVEST’s new point of purchase display seems a fairly normal sportswear display rack. There’s the large base with branding, a vertical bar for hanged garments and a colorful assortment of SCOTTeVEST’s gadget-friendly activewear. When a customer approaches, however, the seemingly normal automatically activates. A locally targeted video is launched while a biometric scanner estimates the customer’s size and begins guiding the purchase process. After the customer tries on a garment, a hangtag can be scanned to activate a rundown of the features, and if there are any questions, the PoP display will open a live video chat with a SCOTTeVEST customer service rep. The eventual purchase can be rung through a standard payment app on the customer’s smartphone, with the product delivered directly to their door. The retailer gets a commission for the sale, and any online promotions or discounts are automatically pushed to the PoP in each store.
“This is truly the best of both worlds,” said Scott Jordan, SCOTTeVEST CEO and founder, “a hands-on truly awesome product experience plus online fulfillment – never lose a sale if an item is locally out of stock.”
The “silent salesperson” has found its voice, plus a lot more.
Yet as intriguing and potentially disruptive as SCOTTeVEST’s innovative display truly is, it’s just a tease of the type of smart technologies that are beginning to appear on retail sales floors. At the ever-quickening pace of technology adoption, advanced applications using artificial intelligence, machine learning and natural language processing (sometimes lumped together as AIML-NLP) are creeping into to retail back rooms and sales floors. It’s the type of stuff that regulates current hot technologies in areas such as cloud, mobile, asa-service and social as little more than utility services. And whereas those technology investments up to this point have largely been about chasing the customer across channels, the second wave of technology is more about attacking costs and streamlining operations. It’s a “future of retail” that is much closer than most people think, and, make no mistake, it will dramatically alter a brick-and-mortar business model that is in desperate need of an upgrade.
It’s easier to confine innovations such as robotics, automation and smart machines as primarily being deployed on assembly lines or in warehouses, largely replacing “low-skilled” laborers performing non-thinking, repetitive tasks. Still, more recently we’re also seeing really smart software and machines handle many “white collar” and “thinking-oriented” tasks, such as robo-advising within the financial services sector, reporting on live sporting events, answering and directing customer service calls or taking orders and processing payments at restaurant tables. There is even a free chatbot called DoNotPay that has successfully challenged 160,000 parking tickets – around $4 million worth of fines – in London and New York.
Basically, if your job doesn’t require negotiating, you never have to be clever, and you are not required to personally help others (oh yeah, and you don’t have to fit in tight places), there’s a good chance you could be replaced by smart machines and/or software. At least that’s what researchers in a study led by Oxford University concluded. Out of nine possible traits, those four scored as the most important. Among the many job roles with a high-90s percentage likelihood of being automated were claims adjusters and insurance underwriters; tax preparers, accountants and legal secretaries; real estate brokers and loan officers; payroll and account clerks; restaurant cooks and host/hostesses.
A few retail jobs also were among the most likely to be automated, according to the experts polled by Oxford researchers. A retail salesperson has a 92.3 percent chance of being replaced by machines and artificial intelligence. Cashiers have a 97 percent chance of being automated. That compares to say an advertising sales agent or massage therapist (both 54 percent likelihood) and a choreographer or sales engineer, both at 4 percent likelihood. (Retail managers, meanwhile, had a 28 percent chance of being replaced by machines.)
All told, the Oxford University study found that 50 percent of jobs could get taken over within the next 10 to 20 years – a prediction backed up in a McKinsey report released last year, which suggested today’s technology could feasibly replace 45 percent of jobs right now. Also earlier this year at an annual meeting for the American Association for the Advancement of Science, computer science professor Moshe Vardi proclaimed robots could wipe out half of all jobs currently performed by humans as early as 2030.
A separate study conducted by the White House Council of Economics and introduced to Congress in President Obama’s February economic report examined the chances that automation could threaten jobs based on how much they paid: either less than $20 an hour, between $20 and $40 an hour, or more than $40. The results showed that those making less than $20 per hour were far and away the most threatened. In other words, 62 percent of American jobs may be at risk, according to Bureau of Labor Statistic counts.
Seem farfetched or far away in the future? Well, if someone told you five years ago that driverless cars would soon be on the road, you’d likely say the same thing. And much of the technology that is being used to develop driverless cars is now being applied to the in-store experience.
Meet OSHbot, for example, a 5-foottall retail service robot developed by Lowe’s Innovation Lab in partnership with Fellow Robots. OSHbot has been roaming the aisles of Lowe’s-owned Orchard Supply Hardware in San Jose, Calif., for about a year now. Armed with facial-recognition and the same navigational technology found in driverless cars, OSHbot’s primary functions are to help customers find items on the shelves and help store managers manage inventory. As for the former, the rolling robot will greet a customer in one of seven languages, and sports two informational screens that display upsell opportunities and promotions. Just tell OSHbot what you are looking for, and it will lead you to the location of that item in the store. Don’t know the name of the piece or part you are looking for? Soon OSHbot will be able to take a visual scan of that part and tell the customer if and where it is in the store. All the while, OSHbot can track inventory in real-time, able to tell employees when an item is out of stock, misplaced or has possibly been stolen.
Several organizations, in fact, are marketing data-crunching, machine-learning roving robotic workers to the retail vertical. Bossa Nova Robots, which just scored $14 million in series A funding, says it is ready to immediately deploy robots that collect terabytes of in-store data to help employees keep track of everything on the shelves and can even re-stock them.
“Bossa Nova addresses a multibillion dollar opportunity within the retail marketplace, and is a technology that can be immediately deployed by major retail chains,” said Nic Brathwaite of WRV, which led the round of funding. “The product has already proven successful in stores and integrates seamlessly with existing inventory management systems.”
4D Retail Technology Corp., for its part, recently unveiled its 4D Space Genius, a robotic imaging platform powered by Segway that can scan an entire store in less than an hour. As the robot travels down each aisle, it automatically compiles and processes huge amounts of big data, imaging every product and barcode in every aisle. Once completed, the Space Genius scan provides retailers and manufacturers with the precise location of everything in the store, as well as all instances of price tag discrepancy, missing price tags, empty shelving and more.
The Space Genius also provides retailers with an interactive 3D map of their store, precisely depicting each product exactly as it is displayed on the shelf. This realistic, virtual store can either be displayed on the company’s Web site for consumer use and shopping, or toured remotely by retail executives at headquarters.
“With one click, customers can virtually navigate through any scanned store anywhere in the world and view products on the shelf exactly as they are,” said the company. “As shoppers tour the aisles, they can pull products off the shelf, spin them around to read more product information and add them to a shopping cart to be shipped or delivered by local courier.”
Included with the Space Genius is a 3D planogram application. “Unlike traditional methods of creating planograms, which are based primarily on static, theoretical inputs, the 4D Space Genius intelligently generates dynamic ‘realograms’ based on actual, scanned data,” says 4D.
Then there’s Pepper, a humanoid robot that Pizza Hut expects to have in its restaurants in Singapore by the end of this year. A joint effort with MasterCard and created by Softbank Robotic, Pepper not only will take orders, provide product information and facilitate payment, it will also be able to access customer and sales information in order to make personalized recommendations and offers. And at the Carnegie Mellon University store, visitors will find AndyVision, an autonomous robot that fuses image-processing, machine-learning algorithms, a database of images of the store’s products, a basic map of the store’s layout and navigation sensors to take thorough inventories and tell staff when an item is running low in stock or merchandise is out of place.
“The idea for AndyVision was born out of me being a shopper. I go to a lot of stores and I find it very difficult to find the items I want, and sometimes I leave when I don’t find what I want,” said Priya Narasimhanm, head of the Intel Science and Technology Center for Embedded Computing at CMU.
Best Buy, meanwhile, has begun using Chloe, a robot that retrieves products that customers request from a kiosk, and Target recently began a trial of Tally, a robot that travels through aisles and takes inventory.
RELATED: FloorSpace: An Upside to Markdowns
Some pretty smart machines are being deployed in the outdoor market, as well. Although created largely to enhance e-commerce customer service, The North Face has been working with one of the most intelligent artificial machines, partnering in the development of the first mobile app experience to put Watson, the powerful artificial intelligence computer owned by IBM, to use in a retail environment. The application was designed to help online shoppers pick the perfect jacket for their respective wants and needs without having to bungle through product pages on the small mobile screen. Watson asks questions such as where, when and during which activities the jacket will be used, and then based on the feedback, crunches data and makes a recommendation. During initial trials, says The North Face, users who provided feedback rated the experience a 2.5 out of 3, and 75 percent said they’d use it again.
Osprey Packs likewise recently introduced its Packfinder digital tool which “allows customers to take a step-by-step journey through the decision-making process of finding the perfect pack for the intended activity,” announced the company. Customers answer several simple questions such as the primary activity the pack will be used for, trip length, desired features and price range. Packfinder then determines the best Osprey product solution and creates a report card explaining the selection.
Again, both Packfinder and the TNFWatson tool were created for online sales, but it’s naïve not to notice how the qualification questions and recommendation processes sound a lot like something most outdoor specialty sales staff hear on their first day of training.
It’s also no secret how retailers need to reduce the cost of running physical locations in order to counter a declining percentage of overall sales. It’s pretty safe to assume, after all, that dollars will continue to shift to online, mobile and social. It’s even possible that online sales are just now hitting a critical mass, recently reaching 10 percent of total sales.
It’s not something providers of such technologies are anxious to talk about – most smart technology initially is being marketed as “assisting employees” – but AI-ML-NLP technologies will be deployed to reduce the labor costs involved in keeping physical stores open, if they’re not already having some impact on hiring. (Wal-Mart, for instance, recently said it was six to nine months from beginning to use drones to check warehouse inventories in the United States.)
While it’s certainly true that in many cases AI-ML-NLP technologies empower retail workers and enable stores to deliver an omni-channel experience, in other cases, they specifically handle tasks traditionally executed by retail sales staffs, cashiers and warehouse workers. When the machines are smart enough and the artificial intelligent enough, those tasks are done much more effectively.
That is no slight on the value of retail employees or the customer service they provide. Rather it is an acknowledgement of the rapid advancements taking place across the spectrum of AI-MLNLP technologies, driven most recently by “deep-learning,” whereby large neural networks modeled after the human brain are fed enough data to be trained to do all kinds of things. Such “artificial brains” are the power behind Google’s search, Apple’s Siri, Amazon’s recommendations and Tesla’s self-driving cars. As those advancements increasingly make their way onto the retail sales floor, it becomes almost unfair to make comparisons between human and “artificial” or machine-based capabilities. After all, machines don’t need breaks or vacation days; they’re never late for work, never steal merchandise and can work 14 straight hours, seven days a week without overtime pay or Labor Department disputes.
Sure, even the smartest machines have their limits. There will be maintenance and upgrade costs and break-fix inconveniences, but non-human workforce solutions also have no need for health insurance, worker’s compensation and employment tax, and human employees simply can’t compete in terms of automatically gathering, storing and retrieving on-demand gobs of customer data the way machines increasingly can. Smart machines and robots also can speak multiple languages and be updated constantly with real-time inventory and customer data.
On the other hand, there will always be lots of consumers who prefer the face-to-face of human interaction and real-person problem solving. But there is also most certainly a decent percentage of shoppers who are indifferent or even prefer interaction with non-humans. A recent study by Mintel suggest as much.
Within the relatively high-touch category of cosmetics and beauty products, Mintel found that 45 percent of beauty consumers prefer to search for product information in-store on their mobile devices rather than ask for assistance from a sales associate. What’s more, two in five (39 percent) of those consumers are interested in using, or have used, a store-provided tablet to research beauty products available.
When former McDonald’s USA CEO Ed Rensi recently stated how it would be cheaper to buy a $35,000 robotic arm than hire a $15 an hour employee to cook and bag French fries, it’s was largely said as a claim in the highly charged minimum wage debate. But placing politics aside and looking at the matter purely mathematically, smart machines at those prices can be justified by eliminating the cost of just one full-time employee, especially when factoring in the dollars for training, insurance, sick days, employment tax and so on.
We’re also already seeing “robot-as-a-service” models being discussed, under which the cost to purchase, maintain and upgrade smart and learning machines is lumped into a recurring monthly cost – much like labor. And whereas current retail technology investments in online, mobile, local and social generally need to be justified by a boost in revenue or customer retention, capital for AI-ML-NLP investments may already exist in budgets, shifted over from the labor line item.
Some may say we sound like doomsayers, or at least are inflating the type of hype this publication is usually careful to deflate. Even so, events and advancements that truly disrupt longstanding business models don’t appear very often. When they do, it’s always better to know about then too early rather than too late.
About Inside Outdoor Magazine:
We know outdoor executives have access to plenty of instantaneous information. But one thing decision makers overwhelmingly tell us they can’t get enough of is thoroughly researched and highly scrutinized news and analysis that can be applied directly to improving day-to-day operations.
Having surpassed more than a decade in the industry, Inside Outdoor Magazine’s staff of seasoned outdoor journalists represent years of experience in specialty retail, product manufacturing, business analysis and living the outdoor lifestyle. We believe this mix of backgrounds is a key component to complementing your marketing message with the business modeling, cost-margin analysis, trend forecasting and quantified operational advice that’s already proven to capture loyalty and engage readers, and studies show that reader engagement with media and advertising can drive sales.
Your customers would need to expend significant time and money to amass the information and advice, in addition to, the requisite industry movements, trends and product innovations—that we deliver straight to their mailboxes, for free, all year round.