Artificial intelligence (AI) is gaining considerable popularity with retailers because of its ability to generate a tangible return on their investment. Retailers are applying the power of AI to analyze consumer shopping patterns, control inventory, and make better decisions. AI is the perfect technology for the complex and data rich world of retail.
AI technology has, until now, mostly focused on consumer data to provide recommendations, but more recently, attention is being paid to the importance of how customer buying behaviour can be impacted by the physical location of products, store layout, and how the store is organized.
What are Planograms?
Planograms – the Merchant’s blueprint that determines item selection, shelf location, quantity, and how items are to be visually merchandized – are beginning to leverage the power of AI to optimize the space for increased sales through driving bigger baskets. Historically, analysis, production, and distribution of planograms have been a manual exercise. So, what happens when AI is used to power planograms?
The goal of planograms is to analyze the multitude of data gathered from stores and to determine, for each store, what product should be placed where and at what price to drive the biggest baskets possible. Where AI outperforms traditional approaches is in its ability to rapidly absorb the massive amounts of data required, continuously learn from it, and optimize the best planogram configurations, even in a changing environment such as the one we have seen with COVID-19. The AI then feeds those optimized configurations into software that generates the planograms and puts them in the hands of the stores. Analyzing, maintaining, and adjusting planograms manually is becoming an unrealistic option in today’s complex multichannel world. The analysis required to be competitive and generate planograms that fully maximize the basket size and business results is simply not possible without AI.
How Does AI Optimize Planograms?
How does AI optimize the location? Relying on receipts data, historical product adjacencies and product Halo Effects, AI continuously learns and modifies algorithms to fit and anticipate customer buying behaviour, thereby maximizing basket size. In a simple example, if product A is mainly purchased only when product B is adjacent to it, the AI will learn and adjust. Alternatively, if product C is always purchased with product D, irrespective of whether they are next to each other, the AI will consider alternative more profitable basket building locations. In this example, product D might be relocated to a dead area of the store where it can drive traffic and create impulse or associated sales. This seems simple in theory, but billions of potential configurations must be considered based on the thousands of skus, all the potential locations, the impact of seasonality, the stores physical layout, shelf holding power, etc. The complexity is far beyond what people and regular big data analytics are capable of. It is only possible through the development of AI.
Benefits of Using AI to Create Planograms
A major complementary benefit of using AI to creating planograms is that it will be directly tied to forecasting. Accurately predicting demand is vitally important for an efficient supply chain: ensuring product is in-stock, reducing costly markdowns for the retailer, and providing a more refined feedback loop from which the AI can learn and improve further. Traditional forecasting, and even many AI powered methodologies, don’t include the impact of associated item sales (also known as Halo sales), nor do they account for, or solve, the impact of item location. When retailers use AI to build the planograms, the same AI system also does the forecasting, integrating the impacts of the item Halos and location into its forecast, typically moving the average industry forecasting accuracy of approximately 60% to greater than 90%.
The success that AI is earning with retailers is spreading to all facets of Merchandising, including producing planograms that are a massive step change from how they historically have been produced, and most importantly, how they perform. If AI is done right, its ultimate success will be to not only power planograms but tie all parts of the retail operation together to serve both customers and retailers with intelligent speed. AI is the perfect technology for the complex and data rich world of retail that has not changed all that much in 50 years.