Threshold Fillers Get Returned. Your Free-Shipping Bar Paid for It.
Unico Commerce, FounderThreshold Fillers Get Returned. Your Free-Shipping Bar Paid for It.
The threshold worked. Baskets crossed the free-shipping line, AOV rose, the progress bar glowed full. Then the fillers came back.
Threshold-driven adds are the most return-prone items in most stores: small, cheap, barely wanted, added to unlock shipping rather than to own. AOV counted on day one. The refund, the return label, and the extra pick-pack land weeks later — on an order whose shipping you already gave away.
A threshold that ignores returns is not a growth lever. It is a refund scheduler.
Why fillers return at higher rates
A filler is purchased for the threshold, not the product. Fit was never checked, reviews never read, need never established. The decision quality is the lowest of any add in the cart — closer to an impulse click than a considered choice. Low decision quality predicts high return probability across categories, sharpest in size, shade, and fit-sensitive SKUs.
The economics compound: the filler is usually the cheapest line, so its margin is thinnest; it adds a pick, sometimes a parcel split; and when it returns, the original order often drops back below the threshold it was built to cross — leaving you with free shipping granted on a basket that no longer qualifies for it.
Rules for fillers that survive contact with January
Keep the filler a fraction of the basket. When the add-on approaches half the subtotal, you are not nudging — you are re-planning their purchase. Smaller ratios return less because the core basket still anchors the order.
Only low-return, high-margin fillers earn the slot. Refills, care products, consumables, giftable staples — items with structurally low return rates and real handling margin. Never size-sensitive, shade-sensitive, or fragile SKUs as threshold suggestions, no matter their attach rate. Attach rate without return rate is how margin deaths get promoted.
Curate, don't algorithm-maximize. The recommendation model's third suggestion is usually where the return-prone filler lives. Hand-curate the threshold pool per category and season; let the model rank within a safe pool rather than across the whole catalog.
Price the threshold on net, not gross. For each threshold cohort, track: gross attach, extra shipping cost granted, extra handling, and refunds on filler lines within the return window. The threshold level that maximizes gross AOV is rarely the level that maximizes net contribution. Move the bar to the net optimum, even if gross falls.
Measure per eligible visitor, net of refunds
Holdout at the threshold: same traffic, one group sees the progress bar and filler suggestions, the other checks out clean. Compare contribution per eligible visitor after the return window closes, not attach rate on day one. Report filler return rate separately from core return rate — blending them hides exactly the behavior you need to see.
The takeaway
Thresholds don't raise baskets. They add items — and items added for shipping return for shipping reasons.
Stock the bar with low-return margin carriers, keep fillers small relative to the basket, and score every threshold on net contribution after refunds. Anything else is AOV theater with a return label attached.
Threshold lifting AOV but margin flat? Connect your Shopify store and Unico will curate filler pools by return-adjusted margin and test threshold levels against a holdout — on net, not gross.
