How technology is changing the retail space
Machine learning, used with predictive analytics, is starting to impact the retail space.
Large retailers are starting to utilise machine learning, combined with predictive analytics, to assist them in enhancing consumer engagement and in creating more accurate demand forecasts as they expand into new sales channels. With machine learning, computers learn by mining massive amounts of big data, without human intervention, to provide unprecedented consumer demand insights.
When one couples machine learning with other technologies, like facial recognition solutions, says David Cosgrave, Customer Intelligence Lead at SAS, there is the potential to shift from active engagement to automated engagement in the retail space.
"A good example of how such automated engagement may change the retail experience in the future can be found in the experimental Amazon Go shop. This is a new US store, for Amazon employees, which is completely automated. Consumers simply walk in, choose their products and walk out, and the in-store technology detects who is taking what thanks to facial recognition software. This is combined with deep learning that enables the system to know what products have been taken, and the system automatically debits the user's account."
"Obviously, this is a long way from being rolled out in traditional retail environments, but it offers a glimpse as to how such technology will likely affect the future of retail," he says.
Cosgrave adds that it isn't something that's simply being done 'because they can' either - by offering a counterless, walk-in and walk-out store, the idea is to provide consumers with the opportunity of immediate access to products without the need to wait in queues to pay. This, he explains, could enable bricks and mortar stores to more effectively compete with their online counterparts, as automated stores will have fewer worries in regards to issues like staffing and opening hours.
"Another area where machine learning is coming to the fore is in respect of personalised shopping assistance. Cognitive computing, in association with chat bots, is already starting to automate a number of processes - like checking of account balances - but as the machine learning gets better at understanding language, slang and context, so it will bleed further into the retail environment."
"For example, one local retailer is already testing a solution that's designed to help customers with fashion advice. Using facial recognition technology coupled with big data and analytics, the retailer identifies customers as they walk in and is immediately able to recommend accessories and clothing styles to them, based on its existing knowledge of the client. It is, in effect, a more personalised approach to the shopper."
"Another scenario is one where the customer starts off going to the retailer's Web site and chooses a product, before asking for help to 'complete the look'. The bot then recommends products and accessories, based on its knowledge of the customer. This goes beyond simply suggesting products based on what others have bought - as Amazon does, although this is a good starting point - and could go as far as asking the customer to upload a photo, to demonstrate what the different products and accessories might look like when worn.
This, he adds, brings augmented reality into the process, whereby customers can get an understanding of what the clothing would look like on them, without having to go in-store to try it on physically. Furthermore, once the customer has decided what they want, when they do walk into the physical store, an automated bot would be able to inform a human assistant to fetch the relevant products and bring them to the customer immediately.
Ultimately, the above is a good example of taking the typical technique of product recommendations and, with analytics and cognitive learning, making it both a deeper and more human experience, all at once.
"In the end, machine learning is going to lead to a shift in the retail space from active engagement to automated engagement. We will see technology taking over tasks from information gathering to actual execution. And this entire process will be built on a foundation of solid, advanced analytics," he concludes.