Modernizing Retail Merchandising by Overcoming Data Stupor

Merchandisers are drowning in data but starved for insight. In their work, they must rapidly make decisions on which products to purchase, which to stop buying, those that need a bit of extra promotion, and those that should be moved back in favor of other goods. The days of making these decisions based on a little data and a lot of gut instinct are long gone. Today, merchandisers look at inventory levels, sales forecasts, supplier lead times, historical data and much more just to optimize inventory replenishment. Placement of product in a store or on an e-commerce website requires even more information.

Unfortunately, that data is typically trapped in many different systems. Supply chain and inventory data live in the enterprise resource planning (ERP) platform, while store-level data is usually in point-of-sale (POS) systems, and digital purchase data is in the e-commerce suite. Merchandisers are retail experts, not data analysts. Without extensive integration and analysis work done by IT behind the scenes, data is hard to find, correlate and use. The constant application switching and deluge of siloed information often leaves merchandisers in a data stupor, unable to make sense of it all in time to make quick, critical decisions.

The Limits of Traditional BI

Business intelligence (BI) has helped alleviate the data stupor but hasn’t eliminated it. Of course, modern BI platforms can integrate data from myriad sources using a variety of advanced techniques. A few of the more robust solutions even include rich semantic layers that create a shared understanding of data and business rules making them reusable so that the array of interrelated assets can be managed as a whole rather than one at a time.

This is important because traditional BI relies heavily on a visualization layer to present insights from that data, so organizations need to build merchandiser dashboards and reports — and lots of them. After all, a single dashboard will not contain every insight a merchandiser needs to make decisions, and each merchandiser’s data needs will be different from others.

After decades of dashboard development and a never-ending backlog of new dashboard requests to IT, in many ways traditional BI creates just another “destination” where a merchandiser must go to make daily decisions. It’s clear that traditional BI isn't an answer that can easily flex and scale to accommodate the wide-ranging skill levels and continuously changing information needs of the modern merchandiser.

AI-Powered BI: The Solution to Data Stupor

Thankfully, there's a data stupor solution. By combining generative AI (GenAI) with BI, merchandisers can access all the power of BI without having to rely on dashboards. Instead, they can simply ask for the information they need using plain, everyday language. GenAI on its own lacks grounding in reliable data. After all, it’s based on a large language model (LLM), not data analytics. Its superpower is understanding and creating text, not crunching numbers, so when called on to perform complex analysis it may make errors or simply create them out of thin air. However, when GenAI draws on the precise, accurate, up-to-date analytics contained in the BI platform, merchandisers get the best of both worlds: reliable data that’s accessed via a natural language interface.

By integrating modern, AI-powered BI into their workflows, retail merchandisers can easily ask questions about inventory levels, sales forecasts, supplier lead times, and historical data to optimize inventory replenishment. And, just as important, that data can be presented in whatever form is most useful to merchandisers: a graph, a table, even a summary in ordinary language. Not only does this save merchandisers time, it also reduces carrying costs and allows them to make data-driven decisions quickly, improving operational efficiency, enhancing customer satisfaction, and driving revenue growth.

For example, let’s say a merchandiser is trying to determine whether to stock a new toy from a major manufacturer and they want to understand how similar toys have performed in different geographies. They could simply enter a prompt into a chatbot interface, “Provide me with a graph comparing sales for dollhouses that sold for $100 to $150 last quarter in all of our different regions.” The merchandiser could follow up with additional questions such as, “Break this comparison out according to the SKU of each dollhouse.” Or the combination of GenAI and BI could take the form of an overlay on the web browser so that when the merchandiser hovers the cursor over the toymaker’s name in the CRM system or even a news story, a hypercard pops up with key data, which the merchandiser can inquire about further in a text box: “Show me sales trends for each SKU of this toymaker’s dollhouse products in each region for the last four quarters.”

The merchandiser's data journey doesn’t end at a text prompt. In the near future, AI will be able to anticipate, based on context, a merchandiser’s data needs so it can provide relevant information proactively, without the user having to ask for it. But in the meantime, the ability to ask for exactly the data needed to make a decision will save time, improve accuracy, and push the business forward to greater revenue, profits and growth.

​​PeggySue Werthessen is vice president, go-to-market strategy at MicroStrategy, the world’s first Bitcoin development company.