Nearly every online retailer offers free shipping above a stated order value. That number is not arbitrary, and setting it badly costs money in two different directions.
The threshold is funded by the incremental order
Shipping is a real cost absorbed by the retailer. Offering it free above a minimum is a bet that the customer will add items whose margin covers that cost.
The threshold therefore has to sit above the typical order value. Set at or below it, the retailer gives away shipping on orders that would have happened anyway.
Set far above it, few customers reach it and the offer influences nothing while still costing on the orders that qualify.
Average order value is the wrong anchor alone
Order values are distributed, not clustered at the mean. What matters is how many orders sit just below a candidate threshold, since those are the ones that can be moved.
Retailers examining the distribution look for the point where a meaningful mass of orders sits within one typical item price of qualifying.
That anchoring on item price is why thresholds often sit an unremarkable distance above the mean rather than at a round multiple of it.
Margin determines whether the trade works
A category with thin margins cannot fund shipping from an added item. Heavy or bulky goods make this worse because the shipping cost itself is larger.
Retailers in those categories either set high thresholds, charge flat rates, or build shipping into product prices and advertise free delivery on everything.
The last approach shifts the cost to every customer including those buying small items, which changes price competitiveness on individual products.
Returns change the arithmetic
Items added purely to reach a threshold are returned at higher rates than items the customer intended to buy. The margin funding the shipping can disappear afterward.
If return shipping is also free, a single order can carry outbound cost, return cost and processing cost against one retained item.
This is why categories with high return rates, particularly apparel, treat threshold design and return policy as one decision rather than two.
Presentation affects behavior as much as the number
Showing the remaining amount needed to qualify converts an abstract rule into a specific, small gap the shopper can close.
Recommending items priced near that gap makes the action concrete, which is why the qualifying-item suggestion appears at the cart rather than earlier.
Thresholds also need periodic revisiting. Carrier rates, product mix and prices all move, and a threshold set once against a former cost base slowly stops paying for itself.