45% of Shoppers Are Using AI. Your Inventory Model Isn't
For years, retail demand forecasting worked because consumer behavior was relatively consistent. Shoppers gravitated toward familiar brands, orders fell within predictable size ranges, and historical sales data was a reliable guide for inventory decisions.
A 2026 consumer survey from Locus found that 45 percent of U.S. shoppers now use artificial intelligence as either a primary or secondary shopping tool. That scale of adoption matters because AI is producing several distinct behavioral shifts that are landing directly in fulfillment operations — in brand mix, basket composition, and returns volume.
Retailers that understand where those gaps are opening up are better positioned to close them before they compound.
When AI Picks the Brand, Forecasts Miss
Demand planning systems model brand-level demand based on historical patterns, assuming the brand mix consumers choose from stays relatively stable. AI is disrupting that stability in three ways:
1. AI accelerates brand exploration.
Among AI shoppers in the same Locus survey, 39 percent are more likely to try brands they wouldn't have considered before, more than double the rate of the general consumer base. This means that volume can migrate to brands for which the demand planning system has little historical data. This volume shift isn't always gradual. When an AI assistant consistently surfaces a particular brand for a certain query, it can produce step changes that historical forecasting models have no basis to anticipate.
2. AI encourages larger and smaller baskets simultaneously.
Approximately 37 percent of AI shoppers are more likely to add more items to a single order, while 34 percent feel more confident buying fewer. Both can be true because how consumers use AI varies. Some use it to explore a category and end up with a fuller cart, while others use it to pinpoint exactly what they need and buy only that. Fulfillment infrastructure calibrated to average order sizes produces inefficiency at both ends.
3. AI reshapes returns.
Higher return rates follow directly from brand exploration. As many shoppers begin trying unfamiliar brands, a share of those purchases come back. Sizing uncertainty alone accounts for a meaningful portion. Sixty-two percent of consumers admit to buying multiple sizes with plans to return what doesn't fit, a behavior that compounds when shoppers are navigating a brand they've never purchased from before. Retailers whose reverse logistics operations weren't built for that volume absorb the cost twice: once in fulfillment and again in returns processing.
Brand mix unpredictability, basket size variance, and rising return volume compound each other. Fulfillment infrastructure built around historical averages starts producing stockouts, excess inventory, and picking inefficiencies when all three variables shift at once.
Getting Ahead of AI-Driven Demand
Retailers that are adjusting well to AI-driven demand volatility tend to do a few things differently. For starters, they refresh demand planning cycles frequently enough to catch brand migration before it turns into inventory misalignment. They also model for variance in basket composition rather than optimizing around a historical average. And they build enough flexibility into fulfillment operations to handle orders that fall outside the norm. Increasingly, that means relying on systems that can recognize new demand signals and adjust fulfillment decisions in real time, because the volume and speed of AI-driven demand shifts make manual intervention impractical at scale.
The common thread is treating AI shopping behavior as a permanent input into demand planning. AI-assisted shopping isn't going anywhere and will only grow in popularity over the next few years. The demand patterns retailers are seeing today are the baseline going forward. Retailers still calibrated to pre-AI consumer behavior are essentially forecasting for a shopper that no longer exists.
Russell Hoppes is vice president, solutions and delivery at Locus, a logistics technology company committed to enabling excellence across retail.
Related story: Retail and CPG’s Precision Era: How AI is Reshaping Forecasting, Fulfilment and Customer Engagement
A seasoned supply chain and logistics leader with over 20 years of experience, Russell Hoppes has successfully driven large-scale transformation across complex, enterprise supply chain environments. Prior to joining Locus, Russell spent two decades at UPS, where he gained deep, hands-on experience operating and optimizing one of the world’s most sophisticated logistics networks. At Locus, he leads Solutions and Delivery, working closely with global customers to turn complex logistics challenges into practical, high-impact implementations using automation, AI, and data-driven decision-making.





