Does your free-shipping threshold actually pay?

A free-shipping threshold earns its place when the extra contribution from larger baskets and additional orders exceeds the delivery fees you give up, extra fulfillment costs and test expenses. A higher average order value alone cannot answer that question.

For a store currently charging for delivery, the decision is specific: should you introduce free standard shipping above a chosen basket value?

There is evidence for investigating shipping friction. In its May 19, 2026 article, Baymard identifies additional checkout costs as an abandonment reason and recommends showing the full cost, or a clear estimate, early. That supports making delivery charges visible. It does not establish how much free shipping will improve your store's conversion rate. Baymard's checkout research

The calculation below is editorial analysis. All example figures are hypothetical, not observed results or benchmarks. This is based on current documentation and desk research, not a hands-on product test.

Start with one domestic delivery zone and one standard service. Use recent basket values, product costs and parcel invoices to choose a plausible threshold. Check baskets near that amount individually: an inexpensive extra item might push an order into a heavier parcel band. Our offer-pricing guide explains the underlying contribution-margin calculation.

Use this worksheet for each test group:

Contribution = net merchandise revenue − product costs + delivery fees collected − carrier costs − other variable order costs.

Contribution per visitor = contribution ÷ eligible visitors.

Net test benefit = (test contribution per visitor − control contribution per visitor) × test visitors − incremental test expense.

Net merchandise revenue excludes tax and reflects discounts and refunds. Other variable costs include payment processing, packaging, picking and return handling. Avoid counting refunds twice: if actual refunds already reduce revenue, reserve only for unresolved returns. Fixed overhead remains outside this comparison unless the offer changes it.

Here is a hypothetical test in US dollars. Each group receives 10,000 eligible visitors. The control charges $4 delivery on every order. The offer gives free delivery on merchandise baskets of at least $75 after discounts. Of its 330 orders, 180 qualify and 150 still pay $4.

Worksheet input	Control	$75 threshold
Orders	300	330
Net merchandise revenue per order	$60	$64
Merchandise revenue	$18,000	$21,120
Product costs, assumed at 50% of revenue	$9,000	$10,560
Delivery fees collected	$1,200	$600
Carrier costs, assumed at $8 per order	$2,400	$2,640
Other variable costs, assumed at $3 per order	$900	$990
Contribution	$6,900	$7,530
Contribution per visitor	$0.690	$0.753

The offer generates $630 more contribution. After a hypothetical $200 incremental setup and measurement expense, its net test benefit is $430. This assumes comparable acquisition costs between groups; subtract any difference if traffic costs differ.

The owner should test the offer under these assumptions. The example does not establish that $75 is the best threshold.

Parcel cost is the pressure point. If larger baskets make treatment orders cost $1 more each to ship, the $430 benefit falls to $100. At $2 more, it becomes a $230 loss. The break-even increase is $430 ÷ 330 = approximately $1.30 per treatment order, holding everything else constant. Apply this sensitivity to the difference between groups, not a carrier increase affecting both equally.

Before launch, check that the advertised threshold matches checkout behavior. Shopify's free-shipping discount documentation says qualifying product values use discounted prices and exclude taxes. It also documents country restrictions and exclusions for shipping rates above a specified amount. These are current documented controls, not a newly announced October change; the page shows no publication date. Shopify's free-shipping documentation

Keep a decision record containing the threshold, delivery zone, visitor assignment, orders, revenue, product costs, collected shipping fees, carrier invoices, return allowance and test expense. Include every assigned eligible visitor, even those who never build a qualifying basket.

Randomly assign visitors where feasible and keep their treatment consistent. Hold advertising, other discounts and delivery promises steady. Choose the sample requirement and maximum affordable loss before launch; do not declare victory after a few extra orders. Allow returns to mature before the final calculation.

Revert if the agreed loss limit is reached. At the planned review, keep the threshold only if the contribution gain covers its costs with enough allowance for uncertain returns and parcel charges. If the result remains inconclusive, retain the existing delivery policy and clearer upfront pricing. Larger baskets are useful only when enough money survives fulfilling them.