Google LLC says retailers can leverage its newest artificial intelligence service to get a better handle on inventory management, streamline their logistics operations, boost business planning and improve overall customer satisfaction.

Announced today, Vertex AI Forecast is a new service that helps retailers get closer to real-time forecasting by accurately predicting demand for the various goods and products they sell.

Craig Wiley, director of product management for Google Cloud AI, explained in a blog post that accurate forecasting is critical for retailers to ensure they have the right products available in the right volumes and at the right locations.

For retailers, inventory management is a delicate balancing act. Customers don’t want to see items out of stock, but at the same time, too much inventory is wasteful and costly. Wiley points to data from IHL Group that suggests retailers lose more than a trillion dollars per year as a result of inventory mismanagement. However, a study by McKinsey & Co. shows how a 10% to 20% improvement in demand forecasting accuracy can lead to a 5% reduction in inventory costs and a 2% to 3% boost in revenue.

That’s not all, though, for demand forecasting can also help retailers to staff their stores and support centers more efficiently by accurately predicting customer levels. It can also help with planning promotions and evaluating the factors that impact both physical and online traffic.

The challenge for retailers is that as they expand to new locations and add more products, their data becomes vastly more complex. “Unconstrained activities through the pandemic have only accentuated supply chain bottlenecks and forecasting challenges as the pace of change has been so rapid,” Wiley said.

Vertex AI Forecast is meant to help retailers overcome those challenges. It’s a new addition to Google’s Vertex AI machine learning operations platform that debuted last year, giving retailers the ability to ingest datasets of up to 100 million rows from BigQuery or standard CSV files. Then, it evaluates hundreds of different machine learning models using its model architecture search feature to find the one that performs best with that dataset.

Once the best machine learning model has been found, Vertex AI Forecast is able to generate a “highly accurate forecast” that works on multiple levels, tying product demand for individual items at the store or regional level. Retailers thus gain the insights they need to ensure they don’t run out of stock. In addition, Vertex AI can ingest unstructured data to understand how factors such as the weather, freight charges, commodity prices, shipping costs and more can affect demand.

“Such attribution applies not only for the overall forecast but for each individual item at every point,” Wiley explained. “For instance, promotions may be the main drivers of demand in the clothing category on weekdays, but not during the holidays. These kinds of insights are invaluable when decisions are made on how to act on forecasts.”

Early adopters of Vertex AI Forecast say they’re already seeing good results from the more accurate demand forecasting it provides.

Amaresh Siva, senior vice president of innovation, data and supply chain at Lowe’s Cos. Inc., said his firm operates more than 1,700 stores across the U.S., so the need for highly accurate forecasting is critical.

“Using Vertex AI Forecasting, Lowe’s has been able to create accurate hierarchical models that balance between SKU and store-level forecasts,” Siva said. “These models take into account our store-level, SKU-level and region-level inventory, promotions data and multiple other signals, and are yielding more accurate forecasts.”

Photo: gonghuimin468/Pixabay