NEW YORK — As soon as you approach Pepper, a four-foot-tall robot, she starts sizing you up.
Thanks to facial recognition capabilities, Pepper can determine your gender and age bracket. And as you begin asking her questions, she can draw from a vast volume of cloud-based information to give what she thinks are relevant answers. If you smile, she can tell the conversation is going well and that you’re finding her answers helpful. If you don’t, she might ask you if she’s misunderstanding your requests.
Pepper’s maker, Softbank Robotics, has a vision of a world in which many retailers incorporate this technology into brick-and-mortar stores, in which it feels normal and reflexive for you to approach a robot with customer service questions.
It’s part of a wider push across the retail industry to bring more automation and data science to one of the few parts of the business that largely remains an art: the act of making a sale.
At the National Retail Federation’s Big Show, an event here attended by tens of thousands of industry professionals, demonstrations abounded of technology that could assist a store employee in closing the deal — or, in some cases, that could answer the very questions clerks might typically help with.
These innovations present tantalizing possibilities for retailers, who can ill afford to lose any opportunity to satisfy or even upsell a customer at a time when mall foot traffic is tumbling precipitously. But the advances also raise thorny questions about what the retail workforce of the future will look like if a growing array of tasks can soon be punted to robots or tablet applications.
When you first encounter Pepper, it’s hard not to be struck by the quirky novelty of the situation: You’re being chatted up by something that looks part animé cartoon, part “Star Wars” stormtrooper. And yet the interaction feels strangely familiar, because of how uncannily humanoid her gestures are. From the way she cocks her head when asking a question to the way her fingers curl up when she draws her hands to her hips, it all feels very integrated.
Because of that, “one of the challenges we have from a marketing perspective is managing your expectations of what a robot like this can do,” said Steve Carlin, vice president at Softbank Robotics.
In other words, the form-factor is so compelling that people start asking Pepper questions that, at least for now, are out of her depth.
Demonstrators offered several possible use cases for Pepper: Retailers could position her at an endcap — the eye-catching display at the head of an aisle — or in a specialty department, where she could answer questions about a featured product. Or she could help you identify which shoes you might want to buy based on what your priorities are. On a budget? She’ll steer you toward the cheapest pair. Like to be comfortable? She’ll suggest a functional pair of sneakers.
After being used at Softbank Mobile stores in Japan for several years, Pepper made her debut in the U.S. market in November, launching at two Westfield Malls in California. In that iteration, she is programmed to do things like give shoppers directions to a certain store.
Don’t expect to start seeing a massive army of Peppers right away: Softbank says that there is only one facility in the United States that is currently equipped to service and repair the robots, and the company doesn’t want to deploy the technology too widely until it has built out more logistics and capacity for managing that.
Plus, they don’t think they’ve come close yet to unlocking the full range of what Pepper is capable of, as software developers are still being unleashed to write programs for the technology. Think, for example, about how you initially came to understand the smartphone’s power: It was largely thanks to apps built for those devices.
Retailers may be more apt to get on the bandwagon when they can see more use cases. So might hotels, cruise ships, airports or other places that rely on customer service.
Pepper’s makers stress that they don’t see the technology as something that replaces human workers. Rather, they see it as a supplement. While Pepper is answering simple questions such as, “Where’s the restroom?,” the theory goes, a sales associate can be handling complex tasks such as recommending shoes to go with your dress, or helping you find the last pair of jeans in your size.
The idea of technology as a support system for a sales clerk could be seen at other displays during NRF’s trade show. Toshiba, for example, showed off something it calls Lift ’n Learn. When a shopper lifts an item off a shelf, it triggers more information about that product to come up on a large screen behind the display. But certain customer actions can also send a notification to a sales associate that someone is interacting with that display, providing an opportunity for an employee to swoop in, recommend related products, and so on.
Findmine is a start-up that uses artificial intelligence to curate complete outfits from a retailer’s current merchandise assortment. (You may have seen the technology on e-commerce sites such as John Varvatos, where it powers a “complete this look” module that shows you a sweater, shoes and jacket that would look good with the jeans you clicked on.) Findmine has the ability to deploy this technology in physical stores, too. So, it could be incorporated into tablets that associates tote around the store. Then, when they notice you are really loving a certain button-down shirt, the employee could use the technology to make recommendations to you about a full ensemble that is consistent with the brand’s aesthetic.
There are other cases where a worker wouldn’t have to step in at all. Philips Lighting demonstrated a technology called visible light communications, or VLC, in which store lighting can be used to pinpoint your precise positioning and facing within an aisle of a store. (For this to work, you have to have your smartphone out of your pocket and have opted in to allowing the retailer’s app to locate you.)
A spokesman offered an example of how this might be used in a pharmacy for what he called “selective selling”: If you’re hovering around the flu medicine, maybe the retailer’s app could also nudge you to also buy tissues and cough drops.
None of these technologies threatens to completely replace sales associates. But collectively, they might either chip away at the value of those kinds of roles, or at very least alter the skills they call for. Either way, it suggests we’re on the cusp of major change in what it means to be on the front lines at the mall, big-box store or supermarket.