Testing a Free Shipping Progress Bar in the Cart

Order value data with carts piling up just below the free-shipping threshold, or surveys and support tickets showing that shoppers do not know the condition, are the signals that justify this test. When the bar works, carts near the threshold get more add-to-carts and more orders land above it, while order completion holds. If shoppers know the threshold and still leave at the shipping fee, the fee itself may be the problem, so look at that first. If the threshold or promotion terms will change during the test, postpone it, since you cannot isolate the bar's effect while the threshold moves.

Average order value (AOV)

Compare revenue per completed order. Because AOV includes buyers only, read it alongside order completion and revenue per visitor.

Also track

• Add-to-cart rateCheck whether seeing the free-shipping threshold changes add-to-cart.
• Order completion rateTrack order completion among assigned visitors as cart value changes.
• Revenue per visitor (RPV)Compare revenue per assigned visitor, whether or not they purchase.

• ChangeAdd a progress bar showing the amount remaining to reach free shipping in the variant only.
• Keep constantKeep the shipping threshold, product prices, promotions and message logic the same in both arms.
• Cart behaviorTrack both items added and removed to understand how the threshold message affects cart changes.
• Orders and revenueRead AOV alongside order completion and revenue per visitor.
• DeviceCheck that the bar renders correctly and does not obstruct the cart flow on mobile or desktop.

1. 1An AOV gain counts only if order completion and revenue per visitor do not fall. If either drops, larger carts may be coming from fewer orders, so do not call the result a success.
2. 2This finding is tied to this threshold, message and placement; test color, message or position changes separately. Check that the message updates correctly once the cart reaches or passes the threshold. Place the bar so it neither covers the main CTA nor pushes it down; otherwise the result reflects the placement rather than the bar itself.
3. 3When AOV barely moves, look at items added and removed together; the threshold message may lead some shoppers to add items and others to remove them, and the two can cancel out. Before drawing conclusions, confirm that each arm received its planned share of visitors (sample ratio mismatch). The A/B test significance calculator shows whether the add-to-cart rate gap could be due to chance; AOV is an average it does not test, so read that gap with the same caution.