Can robots rescue retail?
The retail sector is at a crossroads. Hit by structural changes as spending shifts online and an uncertain economic environment as well as fast shifting consumer tastes and trends, many in the industry are really struggling. Could robots and artificial intelligence (AI) come to the rescue?
Excitingly, AI is already being deployed in retail, where pure-play online retailers have an advantage in data collection and brick and mortar players can also see healthy returns on investment.
Early and enthusiastic adopters include European online fashion platform Zalando, which has used so-called deep learning to improve warehouse efficiency and in fraud detection and the personalisation of the shopping experience, as well as London-listed ASOS (ASC:AIM), which is successfully deploying AI for product, style and size recommendations.
In this article we will take an in-depth look at how AI is transforming the industry and look at some stocks and funds which are exposed to this emerging theme.
With customer expectations constantly evolving, retailers need to adapt business models to futureproof their profits in the face of advanced competition from Amazon, Ebay and many others.
A study from Juniper Research has found that global spending by retailers on AI services will reach a staggering $12bn by 2023, up from an estimated $3.6bn in 2019, so growth in spending of 230%; Juniper expects over 325,000 retailers to adopt AI technology over this period.
Juniper forecasts that retailers will face an AI adoption race, where early mover, AI-equipped retailers will displace slow moving retailers, offering superior service at optimised price points.
According to its research, the use of AI by retailers will unlock efficiencies across back-office operations. Advanced analytics employed in functions such as demand forecasting and automated marketing could make retailers more agile and improve their margins.
Areas of interest to retailers include personalisation, where AI is leveraged to create a more personalised customer experience. The aim is to boost the quality of customer interactions and make a browser more likely to become a buyer. Examples include AI-personalised digital signage, as well as computer vision digital mirrors and an array of visual search technologies.
According to Juniper, in a retail industry that requires more planning than most, ‘of all the different elements of the retail market which AI will have an impact on, demand forecasting and supply chain analytics will be the most vital’.
Juniper believes the number of retailers using AI-enabled demand forecasting will more than treble between 2019 and 2023. ‘This is the bit (of AI) that is here now,’ Juniper analyst Nick Maynard informs Shares. ‘With the rise of collect-in-store and one-off events such as Black Friday, understanding demand and supply chains is more crucial than ever with AI playing the central role.’
What is AI?
Juniper Research defines artificial intelligence or ‘AI’ as ‘a computer program that uses a combination of digital building blocks, such as mathematics, algorithms and data, to solve complex problems normally performed by humans’.
AI has the potential to transform industries ranging from healthcare, finance and transportation to agriculture. It is also proving to be a game-changer in retail, a sector under pressure from higher costs and a surge in customer expectations as digital disruptors raise the bar for personalised service.
Progress is being driven by a subset of machine learning called deep learning, which involves feeding data to neural networks crudely modelled on how we think the human brain might work and using algorithms to have the computer learn from that data.
Deep learning is enabling remarkable improvements in machine translation, natural language recognition and computer vision. At the same time, the proliferation of the internet, particularly on mobile devices, has triggered an explosion in the data necessary to feed deep learning algorithms, and more data is creating smarter machines.
Although the term AI sounds futuristic, we’re already benefiting from this technology in our everyday lives. When Facebook auto-tags your friends in photos, the social network does so through deep learning; when you use Google’s voice search, the natural language recognition is done by deep learning; and when you put your apartment on Airbnb, the recommended rate is generated by deep learning.
Smart checkouts powered by AI technologies such as computer vision could play a key role in the retail sector of the future. ‘Smart checkouts are a little bit further down the track for large scale deployments,’ says Maynard, although he notes ‘checkout apps are becoming very popular, certainly in supermarkets in the UK’.
Maynard highlights visual search as ‘a really interesting one’, pointing out that Swedish fashion powerhouse H&M and Japan’s iconic Uniqlo are among those retailers offering visual search in their apps.
He also flags the ‘virtual mirror’ as an interesting in-store solution. ‘Using computer vision, the virtual mirror can show you what you would look like with say, a certain cosmetic applied, but without you having to apply it,’ explains Maynard.
H&M is an interesting case study, as the clothing colossus is unashamedly investing heavily in artificial intelligence to improve the way it spots trends, plans logistics and attempts to reduce discounts and piles of unsold stock.
Its corporate venture capital arm – H&M CO:LAB – has also invested in London-based e-commerce retailer Thread, which uses AI to power an online personal styling service selling premium-priced fashion.
Ian Shepherd, former retail CEO whose book, Reinventing Retail: The new rules that drive sales and grow profits, is out in June, believes the regular flow of retailers disappearing is a tragedy and is doing his bit to help them succeed.
In his book, he writes extensively about ‘the importance for retailers of gathering customer data, precisely so that they can do the kinds of analysis that AI unlocks.’ Shepherd points out that many of the transactions that traditional brick and mortar retailers generate are essentially anonymous; these retailers use metrics like footfall and average basket size, ‘but these tell you nothing about how actual individual customers behave’.
And this is crucial insists Shepherd. ‘I worked with one business where we proved that 40% of the revenue came from 10% of the customer base. That will be true for many retailers, but if they don’t know who those customers are, they are flying blind.
‘If their most valuable customer suddenly stops buying from them and defects to a competitor, they won’t even notice let alone do anything about it.’ Retailers need to put themselves in a position where they gather information on who their customers are and what they spend.
‘That’s the real business case for loyalty schemes but you can do the same thing by gathering email addresses at the point of purchase, offering Wi-Fi in your stores and many other techniques,’ insists Shepherd.
‘Once you’ve done that, a world of AI and machine learning techniques which have been honed for many years by subscription businesses becomes possible – identify your most valuable customers, work out the “logical next purchase” for each of them, actively manage churn and more.’
RISE OF THE CHATBOTS
Juniper’s research also found the global number of successful retail chatbot interactions will reach 22bn by 2023, up from an estimated 2.6bn in 2019.
Chatbot use by retailers will enable effective automated customer interactions, allowing them to deliver high quality user experiences in a cost-effective way, while boosting customer retention and satisfaction. ‘For me, retail is the space where chatbots make sense,’ continues Maynard, ‘and it is being deployed by retailers in the returns process’.
AI-based chatbots have already become an accepted part of the retail ecosystem, especially in the e-commerce space, where they could help companies improve profitability.
A potentially lucrative revenue source from chatbots is through ‘cart recovery’, where chatbots can remind customers of the products still in their shopping cart and ask them if they are willing to proceed with the purchase, do nothing or clean the cart.
According to Juniper, retailers that don’t adopt chatbots will face strong challenges from more technologically adept disruptors. The research also found that chatbots leveraged for customer service have a strong potential to reduce costs; with deployments realising annual savings for retailers of $439m globally by 2023, up from $7m in 2019.
As Maynard explains: ‘By embracing automated customer service with chatbots, retailers can act in a more flexible and efficient way. The wider retail market means that chatbots are no longer a luxury, they are essential.’
Walmart's AI store of the future
While retail industry disruptor Amazon has been rolling out cashier-less Amazon Go stores, which have shelf sensors that track the products on its shelves, Walmart, the world’s biggest retailer, is also going all-in on AI.
Its website has been redesigned to be more personalised, and the retail behemoth is also testing ways to digitise its brick and mortar stores, having recently opened its so-called ‘Intelligent Retail Lab’ at a Long Island grocery store.
Through digital efforts, ceiling cameras and shelf sensors will enable staff to fix problems and restock items efficiently. Furthermore, the technology will detect when shopping carts run low, spills occur and when shelves need restocking – notifying workers by phone alert when items need to be replaced – while detecting when cash registers need to be opened up to prevent long lines forming and its cameras can even detect the ripeness of bananas from their colour.
When a banana begins to bruise, the cameras send an alert to a worker. Normally, that task would have relied on the subjective assessment of a human, who wouldn’t have time to inspect each bit of fruit.
Five ways to play
1. Investors can gain exposure to the retail AI theme via investment trust Scottish Mortgage (SMT), which has stakes in Amazon, Google-parent Alphabet and Alibaba.
2. Another option is Ocado (OCDO), whose customer orders are picked and packed in highly automated warehouses using swarms of purpose-built robots.
This process makes up part of its Ocado Smart Platform, which is driven by applications of AI and machine learning and is being successfully licensed out to grocery retailers around the world.
3. Within the small cap ranks, Attraqt (ATQT:AIM) provides SaaS solutions that power online shopping experiences. Empowered by machine learning, the company helps retailers to optimise their e-commerce performance, enhancing conversion rates and growing basket values while guiding shoppers to relevant products and content and reducing the retailer’s burden of time-consuming manual tasks.
Last month (8 May), Attraqt agreed to buy Early Birds, a provider of an AI-powered SaaS platform that allows internet retailers to personalise their offering to individual customers across both the online and offline channels.
Among the household names retailers Attraqt has previously helped is Tesco (TSCO). The supermarket implemented Attraqt’s product and accessory recommendations to track and analyse the behaviour of shoppers from items searched to products added to baskets and purchased.
This data enabled Tesco to suggest relevant additional items to shoppers at every point across the user journey, from initial search to point of purchase and in real-time.
4. In November 2018, ASOS announced the global app roll-out of its sizing tool, ASOS Fit Assistant, which uses machine learning to provide bespoke sizing recommendations on its product pages based on previous purchases and returns.
ASOS has been experimenting with machine learning tools for some time; ‘Your Edit’, ‘Style Match’ and the ‘You Might Also Like’ carousel on product pages are all examples of native ASOS tools that use machine learning to improve the customer experience.
ASOS was also one of the first fashion retailers in the UK to launch on Google Assistant, meaning 20-somethings can discover the latest ASOS products across popular categories using just their voice.
5. Mid Wynd International Investment Trust’s (MWY) managers first began looking at the automation theme five or six years ago. As fund manager Simon Edelsten recounts: ‘The heavy industrial growth had already taken place; cars and shipbuilding have been highly automated for years. We were waiting for two things to happen – for the average price of robots to go down, which would increase the range of processes they could be engaged in, and for breakthroughs in their agility.’
Both of these breakthroughs have now happened, explains Edelsten. ‘We have reached the point where machines can pack raspberries and eggs without making an Eton mess in your shopping bag. That will affect retailers.
‘You only have to go into a Tesco on a Saturday now to see how many staff are wandering around the shop loading trolleys for customers who have ordered online. That is an area ripe for automation. But AI and automation can benefit other retailers too.’
Edelsten adds: ‘Many people think that Amazon is killing the high street but there is still life in it. Online clothes retailers still haven’t cracked a way of making sure customers keep all the clothes they order.
‘Returns really eat into profits. That’s where retailers with a physical presence have an advantage. AI and automation are helping them to introduce just-in-time principles to stock control. These “big box” warehouses that you see by the sides of the motorway are increasingly being filled with sophisticated technology that enables retailers to work out what sizes to send where and as soon as the need arises.
‘That should boost sales and also reduce the amount of stock they have to hold, which reduces the floor space you need, cutting costs and boosting margins.’
Edelsten argues the retail winners will almost certainly be those that invest in automation and AI. ‘We don’t take a gamble on picking the right retailer. We have invested instead in the provider of the automation and stock control software.’
A very successful investment for Mid Wynd has been Japanese outfit Daifuku, widely acknowledged as the world leader in warehouse automation and a company now working with Japanese fashion retailer Uniqlo.
Mid Wynd’s other investment addressing the theme is Unibail, which owns the Westfield shopping centre business and trades on a hefty discount to book value because the market thinks retail is dead. Unibail has grasped the fact that retail is a social experience and is stuffing its centres with cinemas and other entertainments to draw shoppers in. Many of the stores renting space are there to provide a human face to the core online offering. Retailers need a physical presence and Unibail provides it.
Edelsten concludes: ‘AI and automation and Amazon – the three As – are seen by many as a threat to traditional retail, but AI and automation can really benefit those retailers that embrace it and the businesses like Daifuku and Unibai on which they depend.’