Free-Shipping Thresholds That Lift AOV But Raise Return Rate
Free-shipping thresholds reliably lift AOV, but the filler item shoppers add to clear the bar often returns at 2-3x your baseline rate — wiping out the contribution-margin gain.
Quick answer
If your free-shipping threshold lifted AOV but blended return rate climbed at the same time, the filler item is the culprit. Model breakeven on contribution margin after returns — not on AOV — and restrict bar-clearing add-ons to low-return categories (accessories, consumables, gift cards) before you scale the threshold higher.
Free-shipping threshold return-rate leak
The pattern where filler items added to clear a free-shipping minimum return at 2-3x baseline, eroding the AOV gain.
A free-shipping threshold ("Free shipping over €75") is one of the highest-ROI AOV levers in e-commerce. The leak is hidden underneath the headline metric: the marginal item a shopper adds purely to clear the bar is bought with low intent, often in the wrong size or colour, and returns at two to three times the rate of the rest of the basket. Once you net out return shipping, restocking, and refunded contribution margin, a threshold that looks like a +14% AOV win can be flat or negative on profit per order.
This page assumes you've already seen the symptom: AOV went up after you launched or raised the threshold, and your returns dashboard ticked up a few weeks later. If you haven't confirmed the link yet, start with the broader AOV up, return rate up margin-leak diagnostic, then come back here for the threshold-specific fix.
Why filler items return at 2-3x baseline
The shopper's job-to-be-done when they hit the cart with €62 of intended purchases and a €75 bar isn't "buy a product I want." It's "spend €13 to save €6.90 of shipping." Intent collapses the moment the math works.
On an apparel store, that filler is typically a basic tee, a pair of socks, or a second size of an item they were already unsure about. Two of those three will be tried, rejected, and shipped back. On a beauty store it's a mini or a shade they wouldn't normally pick — same outcome.
The hidden cost stack
A returned filler item costs you: outbound shipping, return shipping (often free on a free-ship order), pick/pack labour twice, payment-processor fees that don't fully refund, and restocking inspection. On a €15 item that's €8-€12 of variable cost destroyed — sometimes more than the original margin.
How to detect it in your own data
Segment orders into three buckets: under-threshold, just-over (within 15% of the bar), and well-over (>15% above the bar). Calculate return rate, item-level return rate, and contribution margin per order for each.
The leak shows up as a return-rate spike concentrated in the just-over bucket, driven almost entirely by the smallest item in those baskets. If just-over orders return 2-3x more than under-threshold orders, and the returning SKU is consistently the cheapest line, you've confirmed the pattern.
Typical return-rate profile across threshold buckets (apparel, €75 free-ship bar)
| Order bucket | AOV | Blended return rate | Smallest-item return rate | Contribution margin / order |
|---|---|---|---|---|
| Under threshold (<€75) | €52 | 18% | 16% | €14.20 |
| Just over (€75-€86) | €81 | 27% | 44% | €11.80 |
| Well over (>€86) | €118 | 21% | 22% | €22.40 |
How to fix it without killing the AOV lift
You don't have to remove the threshold — you have to control what clears it. The fix is to steer the filler choice toward categories with structurally low return rates, and to model breakeven on post-return contribution margin instead of AOV.
Safe filler categories: gift cards (0% return), consumables and refills (3-6%), accessories that don't have sizing (belts, hair clips, candles, socks in one-size), and add-on samples. Dangerous filler categories: anything with sizing, anything in a colour the shopper hasn't seen, and anything priced at the exact gap to the bar (the dead giveaway of pure filler intent).
The breakeven question to actually ask
Don't ask "did AOV go up?" Ask: contribution margin per 100 sessions, after returns. If your threshold raises AOV 14% but blended return rate goes from 18% to 24%, you need to calculate whether the extra revenue survives the extra refunds — and on most apparel SKUs at typical margins, the breakeven point is around a 5-percentage-point return-rate increase.
Experiments to run this quarter
Test 1: replace the generic "add €X for free shipping" progress bar with a curated cross-sell drawer that only surfaces low-return SKUs sized to close the gap. Measure blended return rate and contribution margin per order, not AOV. A two-week test on a mid-traffic store should reach significance.
Test 2: raise the threshold by 15-20% and measure whether the just-over bucket shrinks (good) or whether shoppers add a second filler (bad). This also surfaces whether your real lever is the threshold height or the filler curation — closely related to the AOV vs LTV trade-off, where short-term basket gains can mask repeat-purchase damage.
Frequently asked questions
Because it was bought to clear a shipping bar, not because the shopper wanted it. Intent at the moment of add-to-cart is the strongest predictor of return rate, and bar-clearing intent is structurally low. The shopper has already mentally written off the item as the price of free shipping.
Usually no. Thresholds remain one of the highest-ROI AOV levers when filler selection is controlled. The fix is to steer which items clear the bar (toward low-return categories) rather than to remove the mechanic entirely. Removal typically costs more in lost AOV than the return-rate leak costs in margin.
Set it at roughly 1.3-1.4x your current AOV. Lower and the lift is trivial; higher and you push too many shoppers into either abandoning or buying a high-return filler. The exact figure depends on your shipping cost and contribution margin per order — model both before testing.
Much less. Beauty SKUs return at 6-10% versus 20-30% for apparel, and consumables return at 2-5%. The leak is most severe on apparel and footwear, moderate on home and accessories, and minor on beauty, food, and refills. Category mix is the single biggest moderator.
In any cart that lands within 15% above the threshold, look at the cheapest line item. If the same handful of SKUs appears disproportionately in the just-over bucket and rarely in the well-over bucket, those are your fillers. Cross-reference their return rate against your category baseline to confirm.
On apparel with ~40% gross margin, a 10% AOV lift is wiped out at roughly a 4-5 percentage point increase in blended return rate. On beauty with 60%+ margins it takes 7-8 points. Run the breakeven on contribution margin per 100 sessions, not on revenue.
Only if your unit economics support it. Site-wide free shipping removes the filler-item leak entirely but absorbs shipping cost on every order — including the small ones a threshold would have filtered out. For most stores under €15M revenue, a well-curated threshold beats site-wide on contribution margin.
Technically yes, practically risky. Carving out specific SKUs from your return policy is a trust signal in the wrong direction and frequently shows up in customer-service complaints. A cleaner fix is to make the only available fillers items shoppers don't return (gift cards, consumables, accessories without sizing).
Filler-item returns predict lower repeat rates: the refund experience itself is a negative touchpoint, and the shopper learns that your threshold is a trap. This is where the AOV-vs-LTV trade-off bites — a basket optimisation that hurts year-two revenue is a net loss even if quarterly margin holds.
Pull 90 days of orders, segment into under/just-over/well-over threshold buckets, and compare return rates plus contribution margin per order. If just-over returns at 1.5x the other two buckets, you have the leak. The diagnosis takes an afternoon in a spreadsheet or one query in your analytics tool.
Get an AI expert review of your site
Paste your URL — Metricuno's AI runs the same heuristic checks a senior CRO consultant would, scoring your page and prioritising the fixes that'll move conversion fastest.