supply-chain efficiencies and improvements are a top customer engagement priority.  As with many priorities today, many retailers are turning to artificial intelligence (AI) and machine learning (ML) to mitigate losses in this area. AI can surface real-time insights into supply and demand, while generative AI (GenAI) presents them in easy-to-understand language. Better demand forecasting with AI Using AI algorithms to analyze historical sales data, seasonal trends and other market factors can help retailers predict future demand with higher accuracy. This means they can more easily adjust orders and stock levels in volatile times to prevent overstocking or stockouts. AI can also analyze stock levels across various locations and recommend optimal inventory distribution, identifying which stores or warehouses might need more stock due to sudden demand increases. In one example, Canadian apparel retailer Aldo Group has created a demand-forecasting machine learning model called Delphine. This model helps Aldo, which specializes in shoes and accessories, better predict demand to ensure it has the right inventory in the right channels. The aim is also to free up staff to focus on more creative tasks, such as predicting fashion trends based on the analysis that Delphine generates. Managing supply-chain unpredictability The Covid-19 pandemic highlighted the vulnerability of trade routes in times of conflict and natural disaster. In such times, shipments get delayed and even lost, shipping costs and storage costs spiral and customer frustration increases when goods fail to arrive. It’s estimated that supply-chain disruptions cost the average company 45 per cent of one year’s profits over a decade. AI-powered platforms greatly increase transparency into supply chains, providing real-time monitoring and alerts that enable retailers to respond much more quickly. By monitoring suppliers, logistics and geopolitical news, potential disruptions can be detected early. Retailers can then make proactive adjustments, such as sourcing from alternative suppliers or rerouting shipments. AI can also assess supplier risk, continuously evaluating the stability and reliability of suppliers based on criteria such as past performance, financial stability and regional risks. These insights help retailers diversify their supplier base to reduce dependency on high-risk vendors. Another strategy with AI is dynamic pricing. In times of supply disruption, AI can adjust pricing and promotions based on product availability and demand trends. This helps manage consumer expectations as well as enhance profitability on limited stock. Using AI for returns optimization Another growing supply-chain issue for retailers is returns. Businesses are getting increased returns every year. In the US, retail returns reached $744 billion last year, with online orders showing a higher rate of returns than in-store purchases (17.6 per cent vs 10 per cent). For clothing and fashion, the return rate rises to 30 per cent. It’s a huge cost for retailers, not only in the value of lost stock – since many items can’t be resold – but also in delivery costs. This is another area where AI can play a critical role. It can prevent and reduce returns, as well as optimise costs in the return process. Retailers can leverage AI to fine-tune the return window based on a retailer’s specific customer base and product mix, aligning customer preferences with cost-control objectives. For prevention upfront, AI and GenAI can improve purchase accuracy, ensuring that consumers are buying products that fit their needs and requirements. AI-powered shopping assistants can help consumers make more accurate purchases, with more detailed and personalized recommendation engines. A GenAI feature that generates product images in a particular color, size or style, or uses augmented reality so a customer can ‘see’ a product in their own home or ‘try on’ an outfit, also leads to a lower chance of buyer’s remorse. Using AI to analyze why consumers return goods is critical to reducing return rates. One reason is bracketing, in which customers deliberately order several sizes and return the ones that don’t fit. Ensuring more consistent size labeling is important, and AI can analyze data on what items are most frequently being returned for being too big or too small. AI is also being used to help reduce organized fraud. By analyzing purchase and return data, retailers can identify items frequently targeted by fraudsters, allowing for more careful screening and tailored return policies to deter fraud. AI can evaluate return histories to detect suspicious behaviors and patterns. This information can then be used to reduce return fraud such as wardrobing – buying an item to wear to a single event and then returning it. Ultimately, AI-driven supply-chain optimization is a strategic imperative for retailers looking to thrive in an increasingly volatile environment. But retailers must invest in technology that enables them to surface, collect and analyze the data they need, and transition to more composable solutions that can better support changing business needs as technology evolves.