The traditional retail model has taken a battering in recent years, largely from internet businesses like Amazon. But new waves of services, anchored to data analytics, machine learning, and cloud platforms, could also empower physical retailers – giving them new tools to compete.
In fact, adopting these tools might be vital to retail’s survival, as some claim the traditional shopping experience is simply ‘broken’.
Kerry Liu, CEO and Co-founder of artificial intelligence company, Rubikloud, suggests that there are three significant components of today’s retail experience that are fundamentally flawed. The first is inventory management, a retailer’s ability to optimise how much to order, how many products to carry in a distribution centre and how much stock to allocate to every store.
“The second big area of retail that we think is fundamentally broken right now is how a retailer processes and promotes its products,” suggests Liu. “The financial analysis behind the pricing, the promotion, the strategy, everything to do with the retail world is very inefficient and can lead to very poor financial results.”
All of which leads to the biggest nail in the traditional retail coffin - the consumer experience. “It’s nowhere near what it needs to be,” says Liu.
So, where does the consumer experience ‘need to be?’ Revamping the front-end of retail starts at the back-end, where cloud computing and analytics are providing new insights into everything from supply chain data to customer behaviour.
Sainsbury’s, for example, purchased the Nectar loyalty scheme for £60 million, touted as a land grab for the vast repository of customer data the system holds. Nectar encompasses over 500 brands including: eBay, British Gas and BP, and Sainsbury’s described the purchase of it as part of its “strategy of knowing its customers better than anyone else.”
But what do you do with all this information? Tesco has 440,000 employees, 6,800 stores around the world and claims to draw in around 80 million shopping trips a week. As part of a push to leverage this customer data and produce solutions based on cloud platforms, the supermarket moved to a data lake model based around the Hadoop framework.
Tesco claims to draw in around 80 million shopping trips a week.
“The outcome of all of this will be to have real time analytics,” says Vidya Laxman, Tesco head of global warehouse and analytics. “That’s what we are working towards.”
You can do some incredible things with real-time data. In the US, Amazon uses data analytics and cloud computing technology at the heart of its Amazon Go convenience store concept. Computer vision, deep learning algorithms, and ‘sensor fusion’ all combine to create ‘just walk out’ technology, removing the need for checkouts.
As you walk in the store, you check in with the Amazon Go app and take whatever you need off the shelves. Whatever you put into your basket gets automatically added to a virtual shopping cart. You can even put things back and the system will recognise the action and deduct the cost of the item you’ve returned, updating your cart automatically. Once you’ve finished, you simply leave the store and, a few moments later, your Amazon account gets charged.
In the UK, online food delivery service Deliveroo uses data to support algorithmic decisions – in which machine learning models are trained and retrained using the most up to date information – and in real-time operational monitoring.
Deliveroo calls this ‘Frank’, a dispatch engine which calculates food preparation times, how long the rider will take to deliver, and arranges the best combination in real time in to get the order delivered in the quickest possible time. These can be amended on the fly by considering travel delays.
But what about the smaller independent stores? NearSt is creating what it calls a ‘real-time local inventory' on the high street and says that its mission is to get more people into high street shops. Its technology creates a view of what products are in stock, in what retailer, in real-time. This is powered by a combination of its own internal platform, NearLIVE, alongside cloud technologies provided by AWS.
As NearSt co-founder Nick Brackenbury explains: “We’re using our technology and data to ensure that whenever you look for a product on a website, app, or service you already use, if it's stocked nearby, you'll see it. We believe the future of shopping is local, and with our NearLIVE technology we're on track to power that future.”
Brackenbury and his team are exploring innovations around two interesting datasets: a view of real-time local inventory in stores and a growing dataset of local consumer demand (where people are searching for what, when).
“By matching real-world, real-time data of product supply (in shops) and consumer demand (with anonymised geolocated searches, browsing, and so on), you can start to do some really interesting things in terms of customer experiences,” Brackenbury says. “What would it mean if we could predict local demand 100 times better than we can today, in real-time, so that the shops near you always have what you want in-stock at the moment you want to search for it? For us, that's an incredibly compelling future for the high street, and one we think is nearer than most of us realise.”
In terms of what a future shopping experience could be like, Rubikloud believes that it will be powered by data analytics, more optimised, more personalised to the customer, and more specialised by industry.
“Our view is that the checkout experience is going to be seamless, so certainly Amazon Go is correct there,” says Kerry Liu. “I think when you go into a store, you’re going to be able to get information about yourself, your shopping habits and more importantly all the information you need about the products you are buying – this will all be available at your fingertips.”
But this future will only be possible by combining multiple data sets – customer information, ecommerce data, app data, inventory data, and supply chain data. “It’s about the consolidation of data sources, new and old, in order to be able to make predictions and solve problems,” Liu adds.
And that’s where big data truly shines.