In a letter to shareholders in October 2017, Jack Ma – the billionaire founder of Chinese online giant Alibaba – coined the phrase “New Retail” to describe the melding together of real world and digital shopping experiences. This wasn’t just blue-sky thinking. Alibaba plans to open 1,000 smart supermarkets in China in the next five years.
Amazon is also launching its first forays into real-world retail. The company is rolling out AI-powered, checkout-free convenience stores. At 72 Spring Street in New York, online reviews decide what goes on the shelves of Amazon’s new store. Until September 2018, the site in lower Manhattan was home to a pop-up lingerie store. Now it hosts Amazon 4-Star, stocking products with a four-star rating or higher, alongside a selection of new and trending items. The shelves have digital displays that update to match online prices, and printed cards feature excerpts from glowing online reviews.
Giants like Amazon and Alibaba have built their empires on convenience. You can pay with your phone, your thumb or your face. Tap an app, hit a Dash button, or bark an order at your AI-enabled voice assistant, and your purchases could arrive within the hour by delivery drone or gig-economy driver. Now, with ventures like 4-star and Fashion AI, these companies are bringing that approach into the real world, eroding the boundaries between online and offline retail.
These tech giants are not alone in building the future of retail. Look at some trainers on Nike’s website, and you will have a pair in your size waiting for you to try on when you visit the real store in Los Angeles. Companies such as Walmart and IKEA are already bringing technologies like augmented reality and robots to the shop floor. From the weekly supermarket visit to one-off impulse buys, the way we shop is changing. These are the trends, companies and ideas that are shaping the future of retail.
The Nike Live concept store acts as a hub for loyal customers. Vending machines give out free products and the shop's stock is determined by data about the local population
Credit Giovanni Solis
Long before Phil Knight founded Nike, he made a living selling running shoes out of the back of his car at athletics meets. In other words, taking the product right to its target audience. Today the company is using data to do something similar with a new chain of Nike Live concept stores. The first opened in July in Melrose in west Los Angeles.
The “Nike by Melrose” store is designed to act as a hub for loyal customers. The store has a geo-fenced area that uses GPS to deliver special offers to people’s phones when they enter. There are also vending machines where members can collect free products: when the store opened, there were long queues to pick up pairs of socks, for example.
The company is opening flagship stores in New York and Shanghai, and is planning to use the Melrose store as a testing ground for a more data-driven approach to choosing what products to market at those locations, as well as for new technologies that could roll out worldwide. It’s also planning to build more neighbourhood stores like this. Next stop: Tokyo.
On average, more than 70 per cent of visitors to a store leave without buying anything. According to the founders of Israeli startup Mystor-E – Asaf Shapira, Roy Drur and Idan Sergi – that’s because static in-store displays aren’t doing a good enough job at persuading people to make a purchase.
Shapira, Drur and Sergi want to bring the online advertising model into physical stores to improve conversion rates, with displays that change depending on who is standing in front of them. Mystore-E’s software can detect a shopper’s gender, age and personal style from a video feed. Stores can combine that with information about external factors such as the weather or current trends to create recommendations tailored to individuals.
Instead of a static mannequin showing the same outfit to everyone, this technology can power digital displays that promote “the right product, to the right customer, at the right time”, says Sergi, the head of strategy.
The company is also incorporating augmented reality into its displays. It recently signed a deal with Signet, a chain of jewellery stores, where customers will be able to try on items in AR, and shop assistants will be sent advice about which products to recommend to each individual based on their characteristics, and what products have done well in the past.
In May 2018, Westfield shopping centre in Stratford, east London, hosted Secrets of the Empire, a four-person virtual reality experience that dropped players into the Star Wars universe, using Magic Leap headsets and haptic vests to provide force feedback. The pop-up experience was developed jointly by VR company The Void, and ILMxLab, a division of Industrial Light and Magic, the visual effects company founded by George Lucas and known for its work on Star Wars, Jurassic Park and Marvel’s superhero movies.
Virtual reality is becoming more common in the concourses of shopping centres, which are looking for ways to increase footfall. But ILMxLab are working on ways to bring interactive digital technologies into the stores too. In February 2018, they created an installation for London Fashion Week in which a hologram of a digital character interacted with real models in real time. According to Mohen Leo, ILMxLab’s director of immersive content, this kind of work could pave the way for branded digital installations that interact with individual customers and respond to their movements.
Alibaba's store uses RFID-tagged hangers to track customer browser
With a team of computer scientists at Hong Kong Polytechnic University, Alibaba has trained an algorithm on images of more than 500,000 different outfits put together by stylists. It’s using the information about what types of clothes work well together to drive smart recommendations for shoppers – your very own AI fashion adviser.
In July, Alibaba opened a temporary Fashion AI concept store on the university’s campus, in collaboration with clothing brand Guess. On arrival, customers scan their smartphone to link their shopping experience with their online Alibaba account. RFID tags and Bluetooth beacons built into the hangars detect which items of clothing they’ve picked up during their visit, so they can walk into the dressing room empty-handed and swipe through everything they’ve looked at on a touchscreen.
If they want to try something on, a tap will alert a member of staff to bring the item to the customer, and the screen will suggest potential accessories, taking into account previous purchases, and what’s in stock at that location. Back home, a Virtual Wardrobe smartphone app allows customers to go back and buy anything that they tried on in store.
So long, squeaky wheels – the humble shopping trolley is in line for a major upgrade. Dash is a robotic shopping trolley that can manoeuvre its way around a supermarket without being pushed. It has a wheel, motors and sensors, and a touchscreen display where customers can input a shopping list or transfer one from their phone.
Dash will then set off, leading the shopper to each item on the list by the most efficient route through the store. The technology, developed by Five Elements Robotics in New Jersey and being tested by US supermarket giant Walmart, allows shoppers to scan their shopping as they go, and pay for it right from the trolley. “This is really a mobile self-checkout,” says CEO Wendy Roberts.
For retailers, Dash offers the chance to retain customers by helping them find products faster, or by cutting long checkout queues. The trolley will even follow shoppers to their car, and then safely return to a docking station inside the store.
Dots on the ZOZOSUIT enable a smartphone app to plot body shape
The Zozosuit isn’t the most flattering garment, but it’s been designed to help everything else in your wardrobe fit better. Users order one for free online, put it on, and then stand in front of their smartphone, rotating on the spot as the app takes 12 pictures of them from different angles.
The tight-fitting black spandex suit, created by Japanese fashion retailer ZOZO, has 350 white dots to create a 3D model of the wearer’s body shape, which the company claims is more accurate than a human tailor.
The measurements can then be used to order custom-fit clothes at a much lower cost. In Japan, a bespoke T-shirt costs in the region of 1,200 yen, about £8. In the first six months after its domestic launch, there were more than a million orders for the Zozosuit, and the service is being rolled out in 72 countries.
Purchasing luxury goods such as designer handbags online is risky. A third of online shoppers say they have unwittingly purchased counterfeit products on the internet. It used to take an expert to spot the fakes, but Entrupy promises to do it with an app and a small handheld microscope.
Entrupy’s lens can zoom in by a factor of 260, revealing features and flaws invisible to the human eye, which are then wirelessly transmitted to a smartphone and compared with a database of images of genuine items, using artificial intelligence.
The US-based company claims the accuracy of its device has reached 98 per cent for brands like Gucci, Chanel and Louis Vuitton, and that fewer than 0.1 per cent of fakes sneak through its authentication process. The company is targeting online second-hand stores and pawn shops. With prices starting at $99 a month, it’s not cheap – but then neither is a genuine Gucci handbag.
Laurence Kemball-Cook founded Pavegen in his bedroom while still an industrial design student at Loughborough University. He was looking for a way to generate clean energy to power street lighting in urban areas, where wind turbines and solar panels don’t work that well.
His solution was to draw energy from footsteps. His invention consisted of floor tiles that can harvest small amounts of power from the downward pressure of people walking over them. The company’s latest version of the Pavegen tile generates three joules per footstep.
When Kemball-Cook installed the technology at his old school he noticed that the kids started running down the corridors. “People really loved walking on it,” he says. Since launching in 2009, Pavegen has broadened its offering. In shopping centres and supermarkets, a new version is used not only to generate energy but to track customers’ movement. “You can make it really fun, increasing dwell time and making people want to come back again and again,” says Kemball-Cook.
The tiles have been linked to interactive lighting displays, and the company has created an app that allows users to see how much energy they’ve generated and convert it into rewards. “We can incentivise people to move from one place to another.” A Pavegen walkway could entice people towards less visited areas of a shopping centre, for example.
Since 2016 the company has also been using data from the tiles to provide an insight into how people behave. Thanks to low-powered Bluetooth transmitters in the tiles, Pavegen can provide retailers with live data feeds that show when customers are visiting their stores, where they’re walking, and whether they linger in front of any particular products or displays. As well as heat maps and patterns of where people are going, Kemball-Cook says the data can be used to conduct sentiment analysis – determining how customers feel in a given moment.
Data can help inform the design of future stores too. For example, Pavegen has found that people prefer to walk next to a window instead of a wall. The technology can quantify how a slight uphill gradient impacts foot traffic, even if people aren’t consciously aware of it.
Today, Pavegen has 35 staff split across a London office in King’s Cross and a research and development campus in Cambridge. Pavegen tiles have been installed at more than 200 locations all over the world, including shopping centres, airports and stores such as the Mercury Mall in east London, Heathrow Terminal 3 and Harrods. The company has worked with brands like Nike, Adidas, Uniqlo, Samsung and Johnnie Walker, and has projects in the pipeline as far afield as California, India and China.
Kemball-Cook has big ambitions: he wants to build “the largest flooring factory the world has ever seen”, to bring the price point for Pavegen tiles – currently £55-£124 a square foot – down to a level where it can cover whole supermarkets. “There’s so much more than just energy here,” he says.
Tommy Hilfiger's interactive mirror suggests items that match what you try on
Mirrors never lie. But at Tommy Hilfiger’s Regent Street store, in central London, they can offer you fashion advice. Dressing rooms feature interactive mirrors equipped with RFID technology. Combined with RFID tags on hangers, this means that the mirrors can recognise items and suggest complementary items when shoppers are trying them on.
With a tap, customers can request alternatives or different sizes and have them brought in. The fashion brand has focused on creating “digital showrooms”, where you can scroll through items and complete orders with interactive touchscreens. The aim is to make the shopping experience more seamless.
That’s also the driving force behind the TommyNow Snap app, launched in 2017. It uses image recognition algorithms to identify Tommy Hilfiger clothing in photographs, and let users buy something they’ve seen in the real world with a few taps. It was unveiled at London Fashion Week, where attendees could instantly buy the outfits being shown on the runway, but it also works with adverts or even on the street.
Eduardo Rivara spent 15 years in South America managing IT projects for the world of e-commerce, using cookies and user accounts to get a good idea of who his customers were.
He realised that in the real world it was much harder for retailers to know who’s walking into their stores, so in 2016 the Argentinian founded Facenote, which uses facial recognition to plug that gap. It enables retailers to identify individuals in the real world, and target them with personalised greetings and offers just as they would online. The technology has been rolled out at retail locations in Chile, Brazil and the US, including clothing store Studio F in Santiago, skincare retailer philosophy in Manhattan, and three branches of Melissa Shoes in Miami. When they arrive at these stores, customers are invited to have their photo taken and submit their name and contact details at a kiosk, or send a selfie from their phone.
The next time they come to the store, a camera at the entrance will recognise them, and display a personalised greeting. It also links that information with a store’s sales database to tell shop assistants who has just entered the store, what they’ve bought in the past, and what they might be interested in. “It’s trying to bring back this personal feeling to shopping,” says Rivara.
Facenote is a bridge between the online and offline worlds – a customer visiting a physical store could be reminded that there’s an item in their online basket, and offered the chance to try it on there and then. If you buy something online and want to collect it in-store, staff could be alerted to get it ready for you the moment you walk in. “There will be less and less friction,” says Rivara.
Rivara made his technology easy for retailers to implement: there is a set-up fee, and then a simple monthly charge per store. Basic feature extraction is done at the store, but the bulk of the processing happens in the cloud. “We don’t require retailers to have any kind of infrastructure in place. They can just use a regular tablet or a regular security camera,” Rivara says.
The company currently has seven staff in Chile and New York, and has attracted interest and investment from marketing firm R/GA and retail accelerator XRC Labs. It has deals in place with fashion and beauty brands in the Americas, including beauty company Coty, and is in preliminary discussions with some larger retailers, including US department store Target.
Rivara is keen to stress that Facenote is entirely opt-in: people will not be tracked unless they have submitted their own photograph. “We focus 100 per cent on the end user, and we believe that customers should be the ones choosing this interaction,” he says. “Face recognition is an amazing technology, but it’s been done wrongly for many years, used by security companies for asset protection and loss prevention. We really believe that using it for better customer service is the way to go. It’s the future of retail.”
At Amazon Go – the online giant’s cashier-free supermarket chain – customers can simply take what they want, and then walk out without paying. The stores are equipped with cameras and computer vision algorithms that detect when customers take an item off a shelf, and automatically charge them when they leave the store.
There are now four Amazon Go locations – three around the company’s headquarters in Seattle, and one in Chicago – but the company’s Chinese rival is way ahead in the race to redefine the supermarket. Alibaba-owned Hema now has 48 supermarkets in 13 cities across the country. In the first four months of 2018, it opened a new store every six days in China, and it plans to hit 1,000 in the next five years.
Hema has drawn comparisons with Whole Foods because of its focus on fresh produce. Each product has a barcode that customers can scan with their phone to find out more about the journey from farm to shelf. Shoppers can choose their own seafood and have it cooked on site at in-store restaurants. At one Hema store in Shanghai, it’s even delivered to your table by a robot.
The chain also uses facial recognition and mobile payments, but where Hema really has the edge over Amazon and western supermarket chains is in home delivery. It can fulfil online orders from addresses within a three-kilometre radius in under half an hour, while most supermarkets require booking a slot at least a day in advance; even Amazon’s service takes two hours. Shop assistants place items ordered online into bags with a unique barcode for each customer, which are then carried to the delivery centre via an overhead conveyor belt.
Handheld scanners and automated kiosks may have killed the checkout worker, but even Amazon Go’s hi-tech cashier-free supermarket still uses actual workers to restock the shelves. But technology has them in its sights.
Retailers lose out on almost a trillion dollars a year because products are out of stock, or customers can’t find a staff member to locate something they want. Walmart is trialling inventory-scanning bots in 50 of its US stores to solve this problem – they use computer vision to check for empty shelves or items in the wrong place, and alert staff.
According to Jeremy King, the supermarket’s chief technology officer, the robots are three times faster than humans, and more accurate. Scout, a robot spun out of work at the University of Pennsylvania by startup COSY, resembles an upright vacuum cleaner. It creates its own 3D models of how a store has actually been stocked, which retailers can then compare to the planograms handed down from head office.
At Lowe’s, the US hardware store that hires science fiction writers to imagine the retail store of the future, customers can ask for help finding products from LoweBot, a touchscreen-enabled pillar that talks directly to users. James Cleveland, COSY’s founder and CEO, says his company’s technology will eventually be used for robots that can restock the shelves too, as well as fulfilling orders placed online. But we will probably always need some human workers – robots aren’t very good at dealing with unexpected items in the bagging area.
IKEA has built much of its success on its unique physical shopping experience – the slow meander through showrooms, the cheap hotdog by the exit, the struggle to get everything into your car. Now the company is also becoming a pioneer in the world of augmented reality.
Its IKEA Place app allows users to place virtual furniture in their home, helping them visualise how it would look in real life. In March 2018 they introduced visual search. Customers can upload a photo of an item in the real world to see similar products from the IKEA catalogue. Further down the line, IKEA is hoping to combine this algorithm with information about a user’s budget, preferences and room dimensions to offer suggestions, in effect creating an AI interior decorator.
AR could eventually even help with the enduring IKEA frustration of actually putting the furniture together. Canadian designer Adam Pickard – who is not affiliated with the company – has created a concept app, called AssembleAR, that walks you through the construction process for that Billy bookcase, overlaying instructions onto the materials in front of you.