What’s on the dinner table this holiday season is part tradition and part taste. And for millions, those tastes are changing rapidly.
Solidly one-third of Americans have stopped or significantly reduced their meat consumption. And it’s not just the explosive rise of specific products like the Impossible Burger that proves consumers, including non-vegans, have a growing appetite for meatless alternatives. Sales of plant-based substitutes for meat, cheese, milk and eggs grew 17% over the past year, according to data from Nielsen and the Good Food Institute.
Meanwhile, every year, searches for vegan Thanksgiving recipes increase. If this trend continues, grocers must imagine a Thanksgiving table without a turkey at the center in order to meet consumers’ growing preference for plant-based alternatives, and carry the specific products they want, or risk losing these shoppers.
Many U.S. grocery stores are already beginning to carry plant-based meat alternatives, and it’s a fast-growing market worth more than 4.8 billion. However, predicting demand for specific products (e.g., Beyond Meat, Impossible Burger, etc.) in this emerging category remains a challenge. Let alone for niche products that are season-specific, such as vegan eggnog.
So how can retailers and grocers offer the goods that consumers want without losing their shirt or risk stocking out? It turns out the answer is one of the best applications of AI in retail.
The Challenge of Predicting Demand for Emerging Categories
Knowing exactly which products to stock and how much inventory to order is one of the most difficult challenges that retailers face. Typically, this time-consuming, manual process involves months of complex demand forecasting and supply chain planning.
Emerging categories add an extra wrinkle to forecasting because demand planners have less data to work with – impacting retailers’ ability to make decisions that carry serious implications for shrink, customer satisfaction and profitability.
If they order too much, retailers are stuck with worthless product left sitting on their shelves. For example, imagine cases of low-volume vegan stuffing that just won’t move, no matter how much they are marked down. Alternatively, if retailers order too little turkey and cannot keep stores stocked, they will miss out on potential sales, as well as frustrate more traditional customers.
So, just a few percentage points can represent millions in revenue opportunity, loss of customer loyalty and reputational risk.
Thankfully, with sophisticated AI modeling, using data from similar and comparable products, retailers and grocers can more accurately predict demand in emerging categories and confidently make informed decisions.
Using AI to Predict Demand for Plant-based Products
With more and more Americans incorporating plant-based foods into their diets, the question becomes, are grocers prepared for a Thanksgiving season that includes less meat?
Those that are integrating AI into their supply chains can most likely answer yes.
By integrating many first and third-party data sources – focusing on factors such as seasonality, retailer sell-thru, promotion performance, cross-product effects and price elasticity – retailers can use AI to calculate improved demand forecasts to support the planning process. They can even compare data on emerging products with expected performance of similar mainstream products and get granular down to the regional or store level to account for differences in taste.
Today’s shoppers want more personalization and products that fulfill their needs. Consumer tastes can vary greatly between markets, and preferences can change quickly to fit the latest trend.
In this competitive market, retailers need to be able to make faster, more intelligent, data-backed decisions. This is where AI can make all the difference in helping executives capitalize on the growing plant-based food market.
-Kerry Liu, CEO of Rubikloud
Kerry Liu is the co-founder and CEO of Rubikloud, the world’s leading AI and machine learning platform for enterprise retailers.