3 Ways Retailers Can Use AI to Be More Sustainable

Sustainability figures into the buying decisions of growing numbers of customers. That’s obvious to everyone from Kroger to H&M. Less obvious has been how artificial intelligence (AI) — poised to reshape many business processes — will play a role in retailers’ ongoing drive for sustainability. It’s early days, but so far AI’s impact on sustainability looks to be greatest in three areas: inventory management, delivery logistics, and emissions management.

AI in Inventory Management

Perishability is a factor throughout retail. Electronics and vehicles get superseded by newer models, fashion tastes change, drugs expire, and so on. But grocers face the greatest challenges here. Some 30 percent to 40 percent of the U.S. food supply goes tossed or composted, appreciable amounts of it in the form of unsold perishables that add up to about 5 million tons a year of U.S. grocery store waste alone.

For grocers, cutting that sort of waste boosts sustainability along with profitability. AI and machine learning is already making a difference here. Take the example of Kaufland. The Germany-based hypermarket chain, part of the Schwartz Group that also owns the the grocer Lidl, tracks an 800-day sales history for each of the roughly 30,000 items, including groceries at each of its 1,200 locations. It then does a 100-day rolling forecast to optimize inventory levels of each product, be it a flat-screen TV or a loaf of bread. Kaufland can use that information to order the right amounts at the right time, as well as discount items that need to sell quickly or go bad. Such precision at such scale isn’t possible without AI.

Clothing and fashion retailers face a different sort of perishability, one where seasonality and fast-changing tastes can turn yesterday’s must-have into today’s landfill fodder. Canada-based Aldo Group has real-time inventory visibility based on customer purchase history and actual and forecast cross-channel demand to engage customers, optimize inventory, and boost sales at its 1,500 stores in more than 100 countries. Going forward, Aldo Group is exploring an AI recommendation engine based on segments, behavior and navigation. Furthermore, the retailer is also considering AI’s applicability in demand forecasting, inventory management and resource allocation. Looking ahead, these systems will be capable of bringing increasingly narrow customer preferences into the inventory management equation, upping the odds of a given item selling rather than being landfilled or recycled. That saves water, energy, time and money.

It's already happening: Today’s inventory management solutions can reduce inventory levels by 30 percent without sacrificing sales, and with no dip on customer satisfaction. Customers are still getting what they want — just not what they don’t want.

AI in Delivery Logistics

There are two main areas in which AI is playing, or promises to soon play, big roles in retail logistics. The first is in transportation management; the second in optimizing a more deeply integrated supply chain.

In transportation management, AI helps route as well as truckload optimization. Store locations and details, such as loading dock locations and the nature of the products being delivered and delivery windows come into play. In grocery, a produce-heavy order from a distribution center will benefit more from an earlier delivery than one heavy on dry goods. Truckload optimization speeds deliveries by packing the truck in reverse delivery order.

The same AI-augmented systems can apply in an integrated supply chain. The challenge here isn’t technical. Rather, it’s about how much data a retailer is willing to share with its suppliers. My sense is that the rise of store brands will force more integration. The vertical integration of store brands sends immediate demand signals straight back to production. To compete with the likes of Costco’s Kirkland Signature brand, which generates more sales than the likes of Nike and Coca-Cola, retailers need to integrate their supply chains to compete on price, react to market signals more nimbly, and better balance inventory with sales.

All that will save fuel/electricity, reduce waste, and improve sustainability metrics.

AI in Emissions Management

The European Climate Law and the EU’s new Corporate Sustainability Due Diligence Directive are two of a growing number of examples of laws pushing retailers to reduce carbon emissions, respect human rights, and protect the environment. These line up nicely with what consumers want from the brands they buy. Getting there means accounting for carbon with a precision and diligence akin to that of financial accounting.

Establishing retail greenhouse gas emissions calls for AI-assisted emission factor mapping to automate, speed up and improve carbon footprint calculations. To return to the Kaufland example, it must establish the embodied energy of all 30,000 products on its many shelves in addition to its own transportation and other emissions. Once the numbers are nailed down, generative AI can then automatically generate environmental, social and governance reports that provide a clear picture of sustainability performance for regulators and the public.

Improving sustainability isn’t easy, and every retailer must walk a unique path. The good news is sustainability and profitability are increasingly intertwined, and AI tools are already helping improve them both.

Dr. Christoph Schroeder works as global vice president and chief solution owner of fashion in the industry business unit retail in the Walldorf office of SAP SE.