Shipping Thresholds
Date: 2026-08-16
The free-delivery minimum. It’s the most-used pricing lever in ecommerce and it’s usually set by copying a competitor — when the right answer is derivable from your own basket distribution in an afternoon.
What it is
A shipping threshold is the order value at which delivery becomes free. It works by giving customers a concrete goal, which lifts AOV — and it costs margin on every order that would have qualified anyway.
Deriving it from your own data
Plot the basket value distribution and find where customers already cluster:
basket value distribution
▌▖
▌▝▖▁▁
▌ ▝▖▁▁▁▁
─┴──┬──┴────┬────────
£38 £50
median ← threshold here
Set it modestly above the median, close enough that the gap is bridgeable with one more item. On the running model — £38 median basket, £50 AOV — a threshold around £50 asks most customers for one small addition.
Too low and you give away delivery on orders that were coming anyway. Too high and it stops motivating — customers give up rather than reach.
The arithmetic
The calculation almost nobody does before setting one:
delivery cost £3.00
contribution per £1 of incremental spend 30%
threshold £50, customer at £38
needs to add £12.00
incremental contribution £3.60
delivery cost absorbed −£3.00
───────
net gain £0.60
Thin. And that’s only for baskets genuinely pushed upward. For everyone already above £50, you’ve simply given away £3.
1,000 orders at threshold £50
620 already above £50 → −£3.00 each = −£1,860
180 pushed up from below → +£0.60 each = +£108
200 below, unchanged → £0 = 0
────────
−£1,752
A threshold can lose money and still raise AOV. Which is why AOV is the wrong metric to judge it on — contribution per order is — Contribution Margin, Average Order Value.
What makes it work anyway
The arithmetic above ignores the conversion effect, which is usually where the value is:
- Unexpected delivery cost at checkout is a leading cause of abandonment. A visible free-shipping threshold removes the surprise, and that conversion gain often exceeds the margin cost
- It’s a competitive expectation in most categories — not offering one costs conversion regardless of the maths
- Padding items are often high-margin accessories, which improves the incremental contribution figure above
So the honest position: the threshold is usually justified by conversion, not by AOV — and it should be judged on total contribution, not on basket size.
Variants
- Tiered — free standard over £X, free express over £Y. Captures more value from larger baskets
- Members only — ties delivery to a loyalty or subscription programme rather than to basket size — Loyalty Programmes
- Category-specific, where bulky low-margin items would otherwise be subsidised
- Progress indicator — “£12 away from free delivery” in the basket. The mechanism that makes any threshold work, and the cheapest part to implement
Testing it
- A genuine A/B test on threshold value, judged on contribution per session rather than AOV or conversion alone
- Watch what people pad with. If baskets reach the threshold via 15%-margin accessories, the threshold is buying volume at the expense of mix — Basket Composition
- Watch return rate. Items added purely to reach a threshold are returned more often, and that cost lands weeks later — Return Rate and Reverse Logistics
- Recompute annually. Delivery costs and basket distributions both move, and a threshold set three years ago is priced against a different business