The secret to successfully using AI to power a new era of retail

The secret to successfully using AI to power a new era of retail

Microsoft’s ShiSh Shridhar explains how retailers are using AI-enabled virtual agents 

Rebecca Gibson |

This article first appeared in the Spring 2018 issue of The Record.

Currys PC World, a subsidiary of UK-based electronics and telecommunications retailer Dixons Carphone, has a new employee named Cami. She’s ‘mildly geeky’, ‘quite confident’ and available 24/7 to help customers find the products that best meet their needs. She can provide detailed answers to any product questions, check stock levels and help customers locate items all within a matter of seconds. Cami isn’t a typical human store associate. In fact, she’s a conversational bot powered by artificial intelligence (AI), the Microsoft Bot Framework and Microsoft Cognitive Services.

Programmed with information from Dixons Carphone’s online buying guide and store colleague training materials, Cami can be accessed via Currys PC World’s website and Facebook Messenger. She currently accepts text-based questions and photos of product shelf labels from customers, and also enables store associates to check stock and access customers’ wish lists.

“It would be impossible for Currys PC World to have associates who can provide 24/7 expert advice on all products at every single store, so customers may leave dissatisfied and empty-handed if they can’t find someone who can help,” says ShiSh Shridhar, Microsoft’s Retail Industry lead for data and analytics. “Cami removes the risk of lost sales and customer churn by quickly providing the detailed information customers want – whenever and wherever they want it. Customers are satisfied because they don’t have to wait for store associates, and shop assistants are happy because they are free to provide a personalised customer experience.”

US department store Macy’s is also using a Microsoft Dynamics 365 AI-powered virtual agent to resolve 25% of customer queries. Available via Macy’s website, the chatbot can also seamlessly transfer customers to a human operator. The ability to do this makes virtual agents particularly effective in retailers’ call centres.

“While the customer is speaking to a human call centre operative, a chatbot will transcribe the call and simultaneously connect to the retailer’s back-end systems to find data related to their query,” Shridhar explains. “The chatbot will surface this information to the live agent in a matter of seconds, allowing them to provide customers with a quick, accurate and personalised response. The technology does all the work in the background, but customers get the all-important personal touch because they’re hearing the information from a human. And, because the call centre operator doesn’t need to manually search for data, they can serve more customers in a shorter space of time.”

Using chatbots and other AI-powered solutions to augment human capabilities is fast becoming a key priority for retailers worldwide. Many are adopting AI systems that empower employees with the data insights and tools they need to work more efficiently, while simultaneously optimising in-store operational processes.

“AI-powered systems are ideal for finding patterns in retailers’ data and combining them with contextual information about weather, socio-­economic trends, geographical location and other factors to unlock insights that help them to make operations more efficient,” remarks Shridhar. “Retailers are using AI solutions to accurately create hyper-local inventories for individual stores, predict optimum product pricing, and to track product demand and stock levels in real time so they can reduce out-of-stocks.”

Europe-based multichannel retailer OTTO, for example, has boosted sales and decreased product returns by using Blue Yonder’s Microsoft Azure Machine Learning-based Price Optimization solution to determine the best product prices according to predicted changes in customer demand throughout the year. US supermarket chain Giant Eagle has deployed Powershelf, which uses sensors to track stock levels and AI algorithms to automatically update products prices based on multiple factors. Meanwhile, US-based home improvement retailer Lowe’s, Japanese electronics retailer Yamada-Denki and San Jose Orchard Supply Hardware Store in California are all using Microsoft Azure and Power BI-enabled robots from Fellow Robots to monitor stock levels on shelves in real time.

“All three of these solutions automate what were previously very time-consuming and resource-intensive processes for retailers, freeing up store associates to focus on engaging in personalised interactions with customers,” says Shridhar. “Plus, Fellow Robots’ autonomous robots can answer customer queries in multiple languages, connect them to human experts via video, or take them to products on the shelves so they’re improving the in-store customer experience in multiple other ways too.”

Other retailers, such as US outlet chain Nordstrom Rack, are using Microsoft cloud-based beacon technology from partners like Footmarks to monitor customer behaviour in their stores.

“By using AI and machine learning systems to analyse customer behaviour data – and combining it with information from other sources – retailers can optimise their workforce so there are always sufficient employees to meet predicted demand on the store floor, at the checkouts or in the stock room,” says Shridhar. “Retailers can also use insights from this data to identify customer preferences so they can capitalise on new opportunities to boost loyalty and drive revenue. If a retailer has a 360-degree understanding of each customer, they can personalise all interactions and ensure that product recommendations, promotional offers and marketing messages are fully relevant to the individuals who receive them. For example, Microsoft has helped website AllRecipes to deploy an AI-driven engine that recommends specific recipes to specific users.”

Shridhar adds that retailers could take personalisation a step further by identifying which products customers have been buying on a regular basis for a long period of time – such as bread or cleaning products – and enable them to buy these items on an automatic subscription basis. “If customers no longer have to buy basic items, they can concentrate on the enjoyable parts of the shopping experience – discovering new products, communicating with store associates on a personal level and trying out the new experiences retailers are offering via mixed reality and other technologies,” he comments.

As more retailers harness AI to help with customer engagement, employee empowerment and optimising operations, Shridhar predicts the shopping experience will become more enjoyable for customers.

“AI is driving many of the things that retailers have being doing for years, but it’s doing them far more efficiently, cost-effectively, quickly and accurately than ever before, transforming the customer and employee experience, and operational processes,” he says. “The beauty of AI is that it’s largely invisible – customers simply notice that their shopping experience is more intuitive, personalised and enjoyable. Meanwhile, retail employees can work more efficiently and serve customers more proactively and personally.”

Subscribe to the Technology Record newsletter

  • ©2024 Tudor Rose. All Rights Reserved. Technology Record is published by Tudor Rose with the support and guidance of Microsoft.