Retailers are already really good at automating their back-office operations and have proven themselves more than keen and willing to invest in technology that will drive productivity and efficiency, especially when eliminating manually-intensive and error-prone human tasks.
They have no problem spending money to save money when it comes to boosting productivity at the checkout with tills to manage transactions, at the shelf-edge with barcodes and scanning systems and in the warehouse using sortation equipment to manage products. In fact, I would argue retailers can’t prove they are buying or selling anything without point-of-sale systems (POS) and merchant acquiring banking services, both of which automate and process increasingly cashless transactions at checkout. Such systems further automate the flow of product data between their suppliers, partners and stores, using financial and enterprise resource planning (ERP) systems, which has given rise to dreaded spreadsheet.
- Cover Story: The business of retail is changing as robots run warehouses and stores
- Case study: Warehouse automation helps Adore Beauty cut costs, packaging and improve delivery times
- Survey: Retailers, complete the Which-50 Automation In Retail survey
Artificial intelligence (AI) developments can afford retailers the opportunity to move beyond automation, to autonomous computing. AI systems aren’t just about digitising manually intensive and error-prone tasks that take place at checkout or during picking or stocktakes. Developing branches of AI, such as machine learning or natural language processing, enable computer systems to make recommendations and potentially act on them with little or no human input or intervention.
Many retail management systems already use sophisticated rules engines and data integration capabilities to route the right data to the right people in retail businesses. But the speed and volume of data being generated now that the average consumer’s shopping experience is influenced by digital, as well as a proliferation of digitally-enabled physical touchpoints, has increased the complexity of managing such large and distributed operations at scale.
Only more autonomous AI-based systems can ensure maximum levels of return on investment when implementing additional customer services that test supply chain efficiency and profitability, such as home delivery or click & collect. These new services challenge retailer requirements to manage much more complex logistics, inventory movements and stock availability and their impact on the planning and forecasting of key operations, including buying, merchandising and marketing when serving both online and offline channels that include stores and customers’ homes.
Again though, AI is not a new frontier. Machine learning algorithms manage increasingly sophisticated ecommerce recommendations engines. A recent example is the new Amazon Go convenience format, which uses advanced “computer vision” AI-based software and camera hardware systems to capture and record a customer’s movements around the store and infer purchase intent from their behaviour. The tech recognises whether customers are putting products in their pockets or bags or not and knows when to charge for payment, enabling a frictionless customer experience. This experience would not be possible in a store full of customers were it not for the autonomous nature of the AI systems developed to support it.
This does not mean the death of the cashier or warehouse picker anytime soon though. There is a long tail to such adoption, and only the largest retailers will feel the need to invest early to maintain their levels of competitiveness and service in the face of increased operational complexity. Nevertheless, those data-driven and digitally enabled retailers who can connect up their view of people (staff and customers) and places (stores, warehouses and delivery points), as well as products at speed and scale using AI-assisted technologies will have an advantage in an age when customers turn increasingly to digital tools such as mobile and voice to blend online with offline and increase the convenience, relevancy and transparency of their shopping trips. Those retailers will also then be able to focus on developing the people it does employ to do more value-added customer facing roles and enhancing their physical stores so they are more of a experiential and community-based destination and rapid fulfilment hub than one for just transacting.
The near-term opportunities for AI lie in optimising core operational areas, like the supply chain, logistics and promotions for most retailers. Then warehouse and store managers, merchandisers, buyers and marketers can manage by exception, using real-time information and computer-generated recommendations taken from multiple internal and external data sources, such as stock, sales and weather, instead of drowning in spreadsheets. An AI-based future will also see less low-skilled, seasonal work, which retailers have historically relied on. Consider that Amazon’s Kiva robotic warehouse fleet constitutes at least 20 per cent of the company’s workforce, performing its roles with varying degrees of AI-enabled autonomy.