As the holiday season approaches, retailers are scrambling to get the right amount of inventory placed at stores and warehouses so they can meet customer demand while avoiding the kind of excess ordering that plagued companies in 2021 and 2022.
Major retailers made significant progress in pairing down inventory levels this year, but the holiday season marks another opportunity for companies to potentially get it wrong, and this is pushing some operators to embrace new technology solutions to inventory management.
- Walmart, for example, is betting on a machine-learning algorithm to predict inventory needs at its 4,700 stores across the country.
- Providers in this space are also getting attention from venture capitalists. Centro, which provides automated inventory control, raked in $2 million in pre-seed funding last week.
The growing interest in cutting-edge tech to help manage inventory comes at a juncture point for retailers. Prior to the pandemic, so-called “just-in-time” supply chains helped companies maintain lean inventories. But this left them without a buffer when supply chains started acting up, which explains why many embraced a “just-in-case” approach in recent years.
Now just-in-time is making a comeback, and technology could be key to making it work.
Feeling the tension: Remington Tonar, co-founder of Cart.com, a third party logistics provider with clients such as Pacsun and Toms, told Retail Brew that there is a “perceived tension between just-in-time or just-in-case inventory approaches” among merchants this holiday season, and that demand forecasting software could help companies avoid overcompensating in either direction.
“One of our core competencies is actually predictive inventory and demand planning,” he said. “All of those tools are available to our logistics customers, and I think a lot of them find extreme value in that.”
Retail news that keeps industry pros in the know
Retail Brew delivers the latest retail industry news and insights surrounding marketing, DTC, and e-commerce to keep leaders and decision-makers up to date.
He explained that the AI and machine-learning powered software uses a mix of historical data and upstream signals such as advertising clickthrough rates and website impressions to better anticipate future demand.
“By leveraging advanced algorithms and data analytics, AI/ML technologies enhance the accuracy of demand forecasting, enabling companies to anticipate market trends, fluctuations, and customer preferences with unprecedented precision,” Jonathan Colehower, global supply chain strategy practice lead at UST, wrote in Harvard Business Review.
- He also referenced a 2022 report from McKinsey & Co, which found that AI-powered inventory control, as opposed to older spreadsheet-based analytic methods, can reduce errors by between 20% and 50%.
Bad data: However, this kind of data forecasting software still faces some serious limits, according to Fabricio Miranda, CEO and co-founder of Flieber, which provides demand planning services.
“More and more companies are starting to understand the power, especially with AI, of having tools to help them out,” he said. “But most of the tools fall short of actually solving the problem.”
He added that the biggest challenge was the lack of reliable and consistent data to feed into the AI-powered algorithms, which means many firms are sticking to manual data-entry.
“There’s a lot of different channels and all of them have different data sources,” he said. “That is the main reason why most people still rely on Excel spreadsheets.”