In recent years, the retail landscape has undergone a profound transformation with the advent of Artificial Intelligence (AI), reshaping traditional paradigms and redefining the customer experience. AI's disruptive impact on the retail arena extends across various facets, from optimizing supply chain management and enhancing personalized marketing strategies to revolutionizing customer service through chatbots and virtual assistants. This technological evolution has not only streamlined operational processes but has also empowered retailers to delve deeper into consumer behavior, enabling more informed decision-making. As AI continues to weave its intricate threads into the fabric of retail, the industry finds itself on the frontier of innovation, adapting to the dynamic demands of an increasingly digital and data-driven world.
Dealing with this level of speed, personalisation, accuracy, and complexity is humanly impossible. So, retailers globally are digitally transforming themselves to be able to use technologies like AI and automation. This is the reason why many of the worlds biggest retailers including Walmart, Amazon, Lowe's, Target, Tesco, Lululemon have big technology centres in India. It's the reason why large Indian retailers are charting a similar course.
Siddhartha Niyogi, CEO of o9 Solutions India, which provides analytics and AI/ML solutions for faster planning and decision-making, says retailers are also using AI/ ML for improving the assortment in stores, to optimise channels of distribution, predict demand, to improve effectiveness of promotions, to optimise discounts.
Ganesan V P, the distribution sector leader at IBM Consulting, India/South Asia, says AI has become a boardroom topic because of the potential it has to solve the challenges in retail. But he says the implementation needs to be thoughtfully done to get the best outcomes.
The first step, he says, is to have a vision and an intent to be clear what are the problems you want to address.
Once that's done, the second step is that you need to ensure you have the necessary data required for AI to learn and train from in the first place. The more good data you have, the better the AI solution will be. Data, Ganesan says, is often the biggest limiting challenge, because the data is in various forms and shapes in different places, and they need to be brought together. Niyogi says while e-commerce tends to be rich in data, its much harder to get data for brick-and-mortar stores. But brick-and-mortar, he says, is today ripe for interventions.
The third step is to identify the best AI or Gen AI model. Todays large language models are built on top of a lot of data. But there are often questions around the quality of that data. So, Ganesan says, you may have to finetune that model, or maybe even build your own model, which can be expensive.
And finally, you need to ensure a good governance framework to ensure, for instance, that personally identifiable information (PII) is not compromised. Ethics is very important in today's day and age. Responsible companies have to lay down the do's and dont's around AI and how they use it, Ganesan says.
