Free shipping statistics: why 41% is not a sales forecast

A free delivery statistic often reaches a webshop without its question attached. A presentation says shoppers want free shipping. Someone turns that into a budget assumption. The shop then pays for transport on orders it would have received anyway, while nobody has checked whether the offer brought enough additional contribution. The number may be accurate and the decision may still be unsupported.

Use original research to decide what deserves a test. Use your own order records to decide what deserves a permanent subsidy. The difference matters most when margins vary between products or when customers qualify for the offer through several different delivery routes. An attractive headline cannot tell you which of those orders can support the cost.

Which free shipping statistics actually measure the same thing? #

They often measure different things. DHL’s 2025 Business Edition surveyed 4,050 businesses across 19 markets: 86% said free delivery and returns improve sales, while 14% offered delivery and returns subscriptions. Its 2026 research covers 29,000 shoppers and 5,800 businesses across 29 countries. The delivery chapter distinguishes shopper incentives from retailer views. These are original publisher findings, not government statistics or controlled estimates of the effect on a particular shop.

Attach the question, population, fieldwork period and commercial sponsor to every number you copy into a planning document. If the source describes an intention, preserve that wording. If it combines delivery and returns, preserve the combination. Otherwise you quietly change the finding before you have used it.

Does a retailer saying an offer helps prove incremental sales? #

It gives you a reason to investigate. It does not supply the counterfactual: what would the same customers have done without the offer? A business may introduce free delivery during a sale, improve its product range and increase advertising at the same time. Its sales report can be correct without isolating which change caused the improvement.

Ask a narrower question for your own shop. Do eligible visitors generate more contribution when they see this specific delivery offer? Keep the product range, stock availability and advertising treatment comparable. Decide whether you are testing unconditional free shipping, a minimum basket rule or a delivery membership. Combining them makes the result harder to interpret.

A membership also changes who receives the subsidy. Its fee, renewal terms and usage pattern belong in a separate model. You cannot use the experience of a frequent subscriber to predict the cost of giving every first-time visitor free transport. Write the eligibility rule before calculating the campaign budget.

What should go in a shareable evidence ledger? #

Keep a short ledger next to the proposed offer. Each row should identify a claim, its original source, the decision it informs and the evidence still missing. This lets a colleague challenge an assumption without arguing over whether free shipping is generally popular. Popularity and profitability are separate entries.

1. 1

   Save the exact population

   Record whether the source covers global shoppers, businesses or a national customer sample. Add the publication year and the date you checked the original page.

2. 2

   State the proposed mechanism

   Write whether the offer should remove a price objection, simplify the total or encourage a larger basket. Select a primary mechanism so the measurement can address it.

3. 3

   List the missing costs

   Include transport, preparation, materials, stock movements and expected return handling. Mark unknown entries instead of silently treating them as zero.

4. 4

   Define the decision rule

   Choose the contribution measure, eligible audience and review window before launch. Include a check for delayed returns and delivery failures.

Give the ledger to whoever owns pricing and whoever packs the orders. They may spot different omissions. Pricing can identify discounted products with little contribution left. Operations can identify baskets that need a larger carton or another service. Both can change the economics without changing the promotional headline.

Which AirPost costs remain when the checkout says free? #

Using the AirPost constants checked on 19 October 2026, the platform charge is €0.85 per shipment. The recommended small-parcel sender fee is €2.25. Together they are €3.10, before the carrier label, materials and other relevant costs. The sender fee is a recommendation; use your actual agreement in the decision.

As an arithmetic unit, ten otherwise identical shipments contain €31.00 of those two components. This is not a reported customer volume or a complete fulfilment quotation. Showing zero delivery charge to the buyer does not remove either component from your order account. It changes how you recover the cost.

AirPost connects webshops with people who hold stock, pack orders and take parcels to a parcel point. Placing stock near customers can change the shipment from an international parcel to a domestic one. Check available services and the actual sender agreement through AirPost for online shops. The network does not establish what your customers will pay or guarantee that a domestic route costs less.

For a complete route comparison, use the European parcel cost guide. For the separate decision about a minimum basket, use the free shipping threshold calculation. Neither replaces the incremental measurement of a new offer.

How can you test the offer without confusing bigger baskets with profit? #

Assign eligible visitors to stable offer groups where your tooling allows it. Keep the assignment through checkout so a returning visitor does not alternate between prices. Record which offer appeared, the available delivery services and whether an order completed. Compare contribution per eligible visitor rather than only the value of completed baskets.

A larger average basket can coexist with a lower total contribution. Some customers may add low-margin products, while existing buyers receive a subsidy they did not need. Include customers who saw the offer and bought nothing. Excluding them changes the question from whether the offer worked to what successful orders looked like.

Let orders mature far enough to include the relevant returns and delivery adjustments. Keep cancellation, partial refund and replacement records connected to the original offer. If you stop at payment, an offer that creates expensive downstream work can appear stronger than it is. Use the same accounting basis in both groups.

If traffic is too limited for a dependable comparison, say so. A short customer interview or an exit question can identify objections, but it cannot provide a precise uplift. Use that evidence to improve the next test. Avoid filling an uncertain revenue forecast with a percentage borrowed from a survey.

When should you keep a paid delivery option? #

Keep it when it offers a meaningful alternative that your shop can fulfil: a different delivery location, a suitable faster service or a clearer appointment. Explain what changes and show the full total before payment. A paid option that merely repeats the free option under another name gives you little useful information about customer preferences.

Review the result by destination and product suitability before expanding the policy. An offer can work for an ordinary domestic parcel and fail for a bulky international basket. Preserve those distinctions in the checkout rule. A global research headline is a starting question; your measured costs and customer choices supply the local answer.