Where should the free shipping threshold be?

The starting point is your current average order value plus roughly the price of one more typical item. Set below your average and you give free shipping to orders you were already getting. Set far above it and the threshold is ignored, because nobody restructures a basket to reach a number they cannot get near.

The useful test is whether a shopper can close the gap with one more decision. If your typical order is one item at a certain price and your threshold requires three, the threshold is decoration.

Once it is live, move it in one direction and measure for a full cycle, ideally four to six weeks so weekly patterns even out. Watch average order value, gross margin per order and total orders together. Raising a threshold usually lifts average order and lowers order count, and the right answer is whichever combination produces more gross profit.

One caution. Do not change the threshold in a peak trading period and then read the result. Seasonal basket behaviour will swamp the effect and you will draw the wrong conclusion from it.

The average hides the shape, and the shape is where the decision lives. Export your last few thousand orders and group them into bands, then look at where the volume actually sits.

Most stores find two or three clusters rather than a smooth curve. A large cluster at the single item price, a smaller one around two items, and a thin tail above that. The threshold belongs just above a cluster, in the gap where a modest addition carries a basket over the line.

Also count the orders that already sit just under your current threshold. A pile of baskets a few pounds short is the clearest sign in the whole dataset that the line is doing work. A flat distribution across the threshold means shoppers are not reacting to it at all, which usually means they never saw it.

Check the distribution by channel too. Paid social baskets are often smaller than search baskets, and email baskets from existing customers are often larger. One threshold has to serve all of them, so set it against the mix you actually have rather than against your best channel.

Free shipping is not free. Every order that crosses the line costs you the real shipping cost, and the extra item that pushed it there only earns you its margin, not its price. So the arithmetic is simple: the gross profit on the added item must exceed the shipping you now absorb, across enough orders to matter.

Work it out on your own numbers. Take the average shipping cost you actually pay on a threshold order. Take the gross margin on a typical add-on item. If the margin on the added item is smaller than the shipping cost, the threshold loses money on every basket it lifts, and the rising average order value on your dashboard is telling you a comforting lie.

This is why the choice of what people add matters as much as the number. Recommending a high margin accessory near the threshold is the difference between a threshold that funds itself and one that does not.

Also account for returns. An item added purely to reach free shipping comes back more often than one chosen on its own, and a return on a free shipping order can cost you twice.

A threshold nobody notices changes nothing, so the message belongs in three places. On the product page, as part of the shipping line. In the cart, as the distance still to go. And in the mini cart, updating live as items are added.

Word it as a gap rather than a rule. Telling someone they are a specific amount away from free delivery is a next step. Telling them free delivery starts at a certain amount is a policy. The first one gets acted on.

Pair the message with a suggestion. One relevant item priced to close the gap, shown right there, converts far better than sending someone back to browse. Choose it by margin and by whether it genuinely goes with what is already in the basket. A single sensible add-on beats a grid of unrelated products.

And once the line is crossed, say so. A short confirmation that delivery is now free is a small piece of reassurance at exactly the point where people decide to stop adding and start paying.

A threshold works when your catalogue lets someone add something sensible. It works badly in three situations, and it is worth recognising them before you spend a quarter tuning a number.

Single product stores with one price point. There is nothing to add, so the threshold either sits below the product price and is meaningless, or above it and requires buying two of something people only need one of. Quantity packs or a subscription are the better lever.

Heavy or bulky goods where shipping is a large share of the order. Absorbing freight on a piece of furniture to lift the basket rarely works. Flat rate delivery, or delivery included in the price, is clearer and safer.

Wide price ranges. If your catalogue spans from small accessories to expensive items, one threshold is too high for half the customers and free money for the other half. Consider a threshold that excludes the highest priced lines, or category level rules if your platform supports them cleanly.

In all three cases the honest option is a simple, predictable shipping price stated early rather than a threshold that quietly costs margin.