If you can take something away from the last two years, it’s the enormous leaps that retail has made in digitization. More or less forced, even small traders now had to deal with this still largely unknown world. But due to this emergency, the last little gear could finally be set in motion to get the entire mechanism around the topic of AI-containing technology going again. And to put the development of the last few years in perspective: the final major technical leap in human history, which can be equated with the present, took place at the time of industrialization.
AI-enabled Technologies Have Been With Us For a Long Time.
What has long been considered a niche solution for large trading companies can now make its way to the standard application with the growing digitalization of small retailers? But similar to digitization, people still need to learn about this complicated, difficult-to-understand application, which may eventually develop a life of its own and wipe out all humanity! This is all science fiction, of course, because the AI-based technologies we are talking about have been around for many years and have evolved to such an extent that they are already being used in everyday retail.
But what exactly do such AI applications look like in retail? Where are they already being used, and how will they be valuable to us in the future? Here is a brief overview of the top 7 used cases of AI technology in retail:
Stores without cashiers
The pioneer of this AI application is the “Amazon Go and Just Walk Out” shopping technology, the functionality of which was only tested under Amazon Personal in 2016 and later made accessible to the public in 2018. The AI-enabled technology records the products that are taken from the shelf and put back again. The money is deducted from the Amazon Go app account if you leave the store with products. This technology hides various AI applications such as computer vision, deep learning algorithms, and sensor fusion. So it is not a single application but an interaction of several AI-containing acquisition programs that communicate with each other and observe what is happening via cameras.
Chatbots as service assistants
One well-established AI-enabled technology is chatbots, which offer an enhanced customer service experience. They not only answer customer questions but also provide support with search queries, information about a new range and present similar products. For example, a chatbot can suggest matching trousers or accessories if a black sweater is ordered. Already 80% of the big brands worldwide use chatbots, including Tommy Hilfiger and Burberry, whose chatbots help customers to navigate through their collections.
In-Store Assistance and Price Matching
The retail industry also invests in technologies that support customers when shopping and relieve staff. One AI-based technology that will get a lot of attention in the coming years is intelligent shelf technology. Here, price tags are exchanged for electronic tags containing cameras and various sensors. If the shelf or perhaps the shelf opposite is emptied due to high demand, the prices are automatically adjusted, and the staff is informed of the vacancy at an early stage. Real-time inventory management that declares war on the archenemy of retail, the empty shelves. The Metro in Poland is currently testing a sustainable version of the principle with a digital two-price system. The aim is to counteract food waste by using an AI application to automatically make the price of products with an expired use-by date cheaper. Time will tell whether less food will ultimately be thrown away.
But the AI-containing innovative shelf technology is also used with a customer-centric focus. Kroger takes it to the extreme with Edge technology. This cooperates with Microsoft’s AI technology, which can run personalized advertising over digital ads with the help of facial recognition. The technology recognizes the gender and age of customers approaching the shelf. If they have activated a release for personalized shop advice via the Kroger app, similar products individually tailored to the customer’s needs (e.g., gluten-free alternatives, low-fat content, sugar-free, etc.) are suggested. If this approval is not granted, generalized suggestions based on gender and age are played.
Supply chain management and logistics
AI-based technologies have also found their place in the logistics sector. Residual and out-of-stock scenarios can be eliminated. With the help of data collection from the sales and demand history and various parameters such as location, trends, weather, or campaign evaluations, AI technologies can calculate exactly which product quantities should be ordered for each period. A prime example of such an application is the Morrisons company in England. With the help of machine learning processes, a degree of automation of over 90% was achieved in store planning in the needs analysis. Based on over 200 factors that were analyzed and processed, the availability of goods could be improved by 30%, resulting in an almost autonomous supply chain.
Machine Learning: Product Categorization
There are already many suitable applications for machine learning, with the classification of product and commodity groups, in particular, benefiting from this AI application for retail. LovetheSales.com is an online shop that uses machine learning to classify over a million products from a wide variety of sellers and sort them into categories for customers. In addition, the computer vision application can recognize the uploaded images of the products, match them, and, if necessary, suggest a corresponding price for sellers.
A significant upswing in retail can already be attributed to visual search systems, also known as visual search. Searching for images on the Internet is becoming increasingly important, but it always requires written input. But you sometimes need to learn the correct terms to describe what you see. An example of this would be plants in the forest, pieces of furniture, or the clothes of your favorite actors. Now visual search should provide a new boost: Instead of entering text for the search term, photos are uploaded themselves. The AI application can use these to search for relevant results and suggest additional suitable products. OTTO has used this application since 2018 via the like/image search for furniture app. Other applications would be Google Lens or Pinterest Lens.
Trend identification of customer needs
This is a link between external and internal data models. An AI application evaluates internal company data (sales, loyalty programs, etc.) and external trend data from search engines or social media to use a holistic exploratory analysis to make trend predictions for the relevant market.