Set a free shipping threshold above your average order, not at it. Orders that would have cleared the line anyway stop paying for delivery, and only the margin on the extra basket earns that back: on an €85 average order at a 38% cost of goods, giving up a €4,90 delivery charge needs a threshold of about €93.
Free delivery is one of the strongest levers in ecommerce and one of the easiest to hand away for nothing. The mistake is almost always the same: the threshold gets set at or just above what a typical basket already is, on the reasonable-sounding logic that customers should not have to stretch far. Half of all orders already clear a typical basket, so the offer immediately discounts them by the delivery charge they were previously happy to pay, and nobody's cart gets any bigger. This post is the arithmetic that puts the line where it belongs, and the two numbers that tell you afterwards whether it worked; the free shipping threshold calculator runs the same arithmetic on your own costs.
In short
- The offer gives away the delivery charge you stop collecting. The parcel was paid either way; what changes is that nothing covers it any more.
- Only the margin on the extra basket pays for that, so the threshold sits well above a typical order rather than at it.
- On the example store's economics an €85 average order needs a €93,02 threshold to be neutral.
- Two numbers settle it afterwards: the share of orders reaching the threshold, and whether the average order value moved at all.
What the offer actually gives away
The parcel was always yours to pay: the carrier charged €6,40 for it before the offer and charges the same after. What changes is the shipping revenue. An order that used to bring €4,90 of delivery income towards that parcel now brings none, and that €4,90 is the whole giveaway on every order that would have cleared the line anyway. Counting the carriage on top of it counts the same parcel twice and pushes the threshold far above where it pays; forgetting the delivery income altogether is how a marketing offer becomes a quiet margin cut.
There is a small mercy in the arithmetic. The gateway charged its percentage on the delivery income too, so giving it up costs slightly less than the full amount, and the calculator handles that fraction rather than pretending it away.
Why the line sits above a typical basket
An extra five euros of basket does not bring five euros of margin. It brings whatever survives the goods and the payment fee, which on typical economics is around three fifths of it, so replacing a €4,90 charge takes roughly €8 of additional spending rather than €4,90. That is the whole reason the honest threshold sits above a typical basket rather than on it.
| Threshold | What it does | Verdict |
|---|---|---|
| €75 | below the average order | discounts orders you already had |
| €85 | at the average order | neutral only if every basket grows |
| €93,02 | the break-even line | pays for itself if baskets reach it |
| €110 | well above | earns, if enough customers stretch |
What the threshold does to the orders under it
The offer changes two populations at once, and only one of them gets discussed. Orders above the line lose their delivery income, which is the cost everyone counts. Orders below it are now looking at a shipping charge that has been made to feel like a penalty, and some share of them will abandon rather than stretch, which is a cost nobody counts because it happens off the order list.
That is an argument for setting the line within reach rather than at an aspirational number. A threshold a customer can clear by adding one ordinary item does work; one that requires them to double their basket mostly converts a cheerful small order into no order at all.
The part the arithmetic cannot tell you
Whether customers actually spend up to the line is an experiment rather than a calculation, and anyone who tells you their number is guessing about your catalogue. What the threshold gives you is the benchmark to measure that experiment against, which is more useful than a prediction because it is yours.
One structural warning: a single threshold across a catalogue subsidises exactly the orders that can least afford it, because a heavy or bulky product costs multiples of a light one to ship. Where your cost per order varies sharply by category, the threshold should too, even though that is more work than a banner. A shipping rate card template laid out by zone and band shows how far apart your own parcels really are before you choose.
Run it properly in a month
- Compute your own break-even threshold from your goods share, your payment fee and the delivery income you currently collect.
- Set the threshold above it, not at it, so the offer has room to earn rather than merely to break even.
- Record two baselines before launching: the share of orders that already clear the threshold, and the average order.
- After four weeks, compare both. If neither moved, the offer is a discount and should be withdrawn rather than defended.
- If your shipping costs differ wildly by product, set the threshold per category rather than sitewide.
The version that usually wins
Stores that do well with free delivery tend to do two things at once: they set the line high enough to matter, and they make it visible in the cart, so the customer knows they are eleven euros away rather than discovering the threshold after the fact. The arithmetic decides where the line goes; the interface decides whether anyone walks toward it. Both halves are cheap, and skipping either is how a good offer ends up costing money.
A correction, made on 1 Oct 2026: an earlier version of this post said the threshold has to pay back both the carriage and the delivery charge. Only the delivery charge changes when shipping becomes free, because the parcel was paid either way, and the €93,02 on this page was always computed on the delivery charge alone.