How Retailers Can Build Trusted, Responsible AI Initiatives

Today’s retailers face no shortage of challenges. Supply chain disruptions, global unrest, inflationary pressures, and a fluctuating labor market make sales difficult for retailers everywhere.

If they haven’t already, retailers should consider the value of artificial intelligence (AI) and machine learning (ML) for support in overcoming these challenges. These technologies deliver advanced insights that can lead to improved customer experiences, more successful promotional strategies, and enhanced operational efficiencies — all of which will positively influence both the bottom line and customer retention.

Whether building upon an already established AI/ML initiative or establishing one from the ground up, here’s how retailers can build trusted, responsible ML models that will deliver unprecedented business value in 2023.

Benefits of AI/ML Tech for Retailers

According to Grand View Research’s latest AI in Retail Market Analysis Report, the global AI market size in retail is expected to expand at a compound annual growth rate (CAGR) exceeding 23.9 percent from 2022 to 2030. Whether it's an AI-powered chatbot answering customer questions about a product or a computer vision solution automating product placement, AI and ML can play an integral role in the retail industry.

ML can deliver actionable insights that enable retailers to gain a better understanding of constantly changing customer behavior and market trends. Retailers can use ML models to accurately estimate future customer demands, increase operational efficiency, lower inventory costs, and adapt retail operations to respond to market shifts — both now and in the future. Historical predictions related to seasonal trends can help retailers prepare for fluctuations in demand. However, AI won’t deliver its full value unless both the business and its customers can fully trust the technology.

When Models Go Wrong

Each year comes with new buying trends and unpredicted changes in consumer behavior, meaning the data from last season can often become outdated. Outdated data in models can lead to disastrous outcomes like biased recommendations and/or inaccurate or incomplete predictions. Some models shift abruptly. For example, the COVID-19 pandemic caused immediate changes in consumer buying patterns that made existing ML models obsolete. Other models might have gradual drift or even seasonal and cyclic drift.

Building Trustworthy AI in 2023

While the onset of 2023 has brought much excitement around the growing adoption of generative AI and large language model (LLM) tools like ChatGPT, a glaring issue remains around its lack of transparency into how these programs arrive at their results. As these tools continue to gain more prominence, especially in the form of chatbots and smart search engines in the retail industry, retailers will need to establish a framework for responsible AI that ensures all model outputs are transparent, accountable, ethical and reliable.

This is where solutions like model performance management (MPM) come into play. Model performance indicates how your model performs before, during and post-production. With MPM, retailers can uncover deep, actionable insights with explanations and root cause analysis, all while giving immediate visibility into ML performance issues to avoid negative business impact. The technology automatically issues alerts on model performance, helping developers respond proactively at the first sign of deviation. MPM allows users to understand and validate the “why” and “how” behind their AI outcomes. Not only will this technology help companies remain compliant with new regulations, but it will also empower them to trust the models they are relying on to make critical business decisions.

The influence of AI systems is rapidly accelerating and it’s more important than ever to ensure that the technology is used ethically and responsibly. The retail brands that can both survive and outpace their competitors in today’s challenging economic landscape will be the ones deploying trusted, responsible AI systems.

Kirti Dewan is the vice president of marketing at Fiddler AI, the pioneer in model performance management (MPM).