Two behaviors are quietly breaking demand forecasting
Warehouse automation firm Locus Robotics surveyed 1,000 U.S. adults on purchase frequency, shopping channels, delivery expectations, and returns behavior, and the results point to two generational habits reshaping retail demand curves. The first is bracketing, where a shopper buys multiple sizes or styles of an item with the explicit intent of keeping one and returning the rest. According to the survey, 61 percent of Gen Z shoppers and 52 percent of millennials have bracketed an apparel purchase, compared with 37 percent of Gen X and only 15 percent of baby boomers. That generational gap means bracketing has become close to a default shopping pattern for the cohorts retailers are trying hardest to win, rather than a fringe behavior.
The second behavior is social-media-driven discovery. The survey found 79 percent of Gen Z and 75 percent of millennial shoppers have purchased an apparel, beauty, or skincare product after discovering it on social media, compared with just 26 percent of baby boomers. More than two-thirds of those buyers completed the purchase within three days of first seeing the product, which is fast enough to create sudden, hard-to-forecast demand spikes as a trend takes off on a platform a retailer may not even be monitoring closely. Together, bracketing and social discovery describe a shopper base whose purchase and return patterns are both more volatile and faster-moving than the models most inventory systems were built around.
Stockouts are costing more sales than slow shipping
The survey's most actionable finding for inventory planners is about what actually drives abandoned purchases. Thirty-three percent of respondents said an unavailable item or size led them to abandon or delay a purchase, the single most commonly cited reason in the survey. By comparison, only 17 percent said they would abandon a purchase if same- or next-day delivery was not available. That ordering matters: retail technology investment over the past several years has leaned heavily toward speeding up delivery, while this data suggests availability itself, having the right size or variant in stock when the customer wants it, is the larger lever on conversion.
Locus frames this as evidence that inventory availability and replenishment speed now matter as much as shipping speed when demand shifts unexpectedly, a reasonable read of the data even accounting for the fact that Locus sells the automation technology meant to solve exactly this problem. The company says it has seen a record number of peak robot requests from retailers responding to these demand surges, though the article did not include the specific volume figures behind that claim. Whether or not a retailer buys into warehouse robotics specifically, the underlying point, that stockouts driven by unpredictable micro-trend demand are a bigger conversion problem than delivery speed, is a useful recalibration for any fulfillment roadmap built mainly around shipping speed.
What 'flex with volume' actually requires
Locus Robotics CEO Rick Faulk put the operational implication directly: 'demand patterns can change in hours and days, not seasons,' and warehouses now need automation 'that can flex with volume, adapt across workflows, and evolve as consumer behavior' changes. That is a sales pitch for flexible robotics, but it also describes a real architectural requirement that extends beyond hardware. A warehouse management system built around seasonal demand planning cycles, weekly batch inventory updates, and fixed labor schedules cannot respond to a product going viral on a Tuesday and selling out of a specific size by Thursday.
For fulfillment technology leaders, the practical question this raises is whether current systems can detect a demand spike in near real time and reallocate inventory, labor, and replenishment priority within that same window, rather than at the next scheduled planning cycle. That capability requires integrating social listening or at least fast-moving sales velocity signals into replenishment logic, not just running a more accurate seasonal forecast. Whether that gets solved through robotics, software-only demand sensing, or a combination, the underlying requirement, closing the gap between a demand signal and an inventory response from weeks to hours, is now a competitive necessity rather than a nice-to-have.
Returns infrastructure has to absorb the bracketing cost too
Bracketing does not just complicate forecasting, it also pushes return volume higher in a predictable, almost built-in way, since a majority of Gen Z and millennial apparel buyers are intentionally ordering items they plan to send back. That has direct implications for reverse logistics capacity: a warehouse and fulfillment operation sized for a lower expected return rate will be chronically under-provisioned for a shopper base where bracketing is closer to a majority behavior than an edge case. Retailers that have not explicitly modeled bracketing into their returns cost assumptions are likely underestimating reverse logistics volume for their younger customer segments.
This also changes the calculus on whether free returns remain sustainable as a customer acquisition tool. If a retailer's growth strategy depends on winning Gen Z and millennial shoppers, and most of those shoppers bracket as a normal purchase habit, then returns processing speed and cost efficiency become as central to the margin model as initial fulfillment cost. Retailers that have invested heavily in outbound fulfillment automation but treated returns processing as a lower priority should revisit that balance given how structurally embedded bracketing now appears to be in younger shoppers' habits.
The roadmap implication
This survey is vendor research, and any retail CTO should read Locus Robotics' framing with the appropriate skepticism about a warehouse automation company telling you that warehouses need more automation. But the underlying behavioral data, generational rates of bracketing and social-driven purchasing, and the finding that stockouts beat slow delivery as a cause of abandoned purchases, describes a real shift that shows up regardless of which automation vendor a retailer ultimately chooses, or whether it chooses one at all.
The concrete roadmap item this suggests is auditing whether current inventory and replenishment systems can react to demand signals within hours rather than days, and whether returns processing capacity has been sized for a bracketing-heavy customer base rather than a lower historical return rate. Retailers whose planning cycles and reverse logistics capacity are still built around last decade's slower, more seasonal demand patterns are going to feel this gap most acutely during exactly the kind of high-volume, high-return holiday period now approaching.
