Testing a free shipping threshold without guessing
Free shipping thresholds can lift average order value or quietly burn margin. How to pick a threshold from your own order data, run the test cleanly, and read the result without fooling yourself.
Start from your order distribution, not a round number
The right threshold sits just above where your orders already cluster. Pull your last few months of orders and find the typical basket value and the spread around it. A threshold slightly above the median order nudges many shoppers to add one more item. A threshold far above it motivates nobody and simply gifts shipping to your largest orders.
Round numbers feel tidy and usually ignore this math. If most orders land in one band and your threshold sits in empty space above it, you have built an offer nobody reaches. Let the data choose the neighborhood, then round to a nearby clean number for display.
What the test is really measuring
Three numbers decide whether a threshold wins. Average order value should rise. Conversion should hold or rise, since a badly set threshold can also make the store feel expensive and depress orders. And shipping cost per order, as a share of revenue, tells you what the lift cost you. Celebrating order value while ignoring the shipping subsidy is how thresholds look brilliant and lose money.
Run the threshold long enough to cover your normal buying cycle, including weekends and a payday pattern if your category has one. A week of data mostly measures novelty. Shoppers also need time to encounter the offer, leave, and return having planned a bigger basket, which is exactly the behavior you are testing for.
Presenting the offer so the test is fair
A threshold nobody notices tests nothing. Show progress in the cart, a clear you are this far from free shipping message with the remaining amount, and keep the message updated live as items are added. That progress bar is not decoration. It is the mechanism by which thresholds change behavior at all.
Keep everything else stable during the test. A threshold launched alongside a sale, a redesign, and a new ad campaign produces a result nobody can attribute. If the number moves, you will not know why, and the business will make a permanent margin decision on unread evidence.
Reading results and deciding
Compare against your baseline period honestly, checking order value, conversion, and shipping share together. A win looks like order value up meaningfully, conversion flat or better, and profit per order up after shipping costs. If order value rises only because heavy orders got free shipping while typical shoppers changed nothing, the threshold is set in the wrong place.
Iterate in small steps. Move the threshold modestly toward the cluster of orders that nearly reached it and test again. Threshold tuning is a dial, not a switch, and the stores that profit from it treat it as an ongoing pricing variable rather than a one time promotion idea.
Related reading
If this topic is useful, these earlier guides go deeper on connected questions: Testing a money back guarantee: framing, placement, and the return math, Testing your homepage hero: product shot, lifestyle scene, or offer first, and Shopify product page tests: what to change one at a time.
Frequently asked questions
How high should a free shipping threshold be?
Just above your typical order value, close enough that adding one modest item reaches it. Pull your order history and set it near the upper edge of your most common basket range rather than picking a round number with no relation to your data.
How long should I test a free shipping threshold?
Long enough to cover your normal buying cycle, usually several weeks rather than days. Short tests mostly capture novelty and weekday skew. You need repeat exposure and a representative mix of traffic before trusting the order value and conversion numbers.
What if average order value rises but profit falls?
Then the threshold is subsidizing orders that would have happened anyway or encouraging unprofitable baskets. Raise the threshold slightly, or pair it with margin aware product recommendations for the add on item, and measure profit per order as the deciding metric.
Should the cart show progress toward free shipping?
Yes. A live remaining amount message is the mechanism that changes behavior. Without it, most shoppers never learn the threshold exists until checkout, which is too late to add an item calmly, so the test would understate what the offer can do.