The rise of the machines (in retail)
Retailers at NRF’s Big Show are hearing a lot about “machine learning.” But what is it and how does it help retailers grow sales?
Machine learning, in which supercomputers learn from mining masses of data on the cloud without human intervention, unlocks insights into consumer demand.
Unlike traditional approaches that rely on time series models, machine learning taps into thousands of data attributes, many of which are publicly available and external to the retailer. Merging external drivers, such as local events, geo-location data, census information, and text descriptions, with a retailer’s proprietary transactional and customer-specific data, the retailer’s understanding of consumer demand improves dramatically.
“Retailers have made do with a variety of time series forecasting tools that came of age when technology was expensive and difficult to use, data were scarce and latent, and growth and profitability were not as dependent as now on fast and accurate ‘granular’ forecasts,” said Greg Girard, program director of worldwide omnichannel retail analytics strategies at IDC Retail Insights. “Analytically astute retailers running complex omnichannel businesses in highly competitive markets must give serious consideration to machine-learning-based forecasting.”
The more granular, local knowledge derived from machine learning drives more accurate forecasts.
“The benefits of machine learning can be enormous for retailers,” said Ron Menich, EVP and Chief Scientist at Predictix. “Every 5% improvement in forecasts can yield a 3% reduction in inventories, with similar increases in sales and margins. With our machine learning-based solutions, improvements of 25% to 50% are possible, particularly at the local level and for the most challenging forecasts, such as promotions and new products where there is no history.”
Some examples of the power of machine learning in retail:
- Fresh food losses are a multibillion dollar a year problem across the grocery industry. Applied to fresh food categories for a leading international grocery chain, Predictix machine learning-based forecasting and supply chain applications have demonstrated the ability to improve fresh food forecasts and reduce wastage by 50%.
- Inventory inefficiency is an $800 billion dollar problem for retailers. Last year alone, a leading U.S. retailer reduced inventories by $80 million with Predictix forecasting applications. The next generation of machine learning-based forecasts from Predictix promises to continue to improve forecasting accuracy and deliver additional value to this retailer.
The rich insights provided by machine learning not only enhance retailers’ understanding of demand to drive sales and inventory effectiveness; they also allow retailers to better shape demand through more effective assortments, prices and promotions.
Consumer sentiment in January hits 11-year high
New York — Consumers are feeling bullish, buoyed by low gas prices and an improving job market. The University of Michigan preliminary consumer sentiment index for January rose to an 11-year high of 98.2, from a final reading of 93.6 for December.
“Gains in employment and incomes as well as declines in gas prices were cited by record numbers of consumers,” stated Richard Curtin, director of the Michigan Survey of Consumers. “More consumers spontaneously cited increases in their household incomes in early January than any time in the past decade.”
The median estimate in a Bloomberg survey of 70 economists projected the Michigan index would increase to 94.1.