The Pricing Science of Free-Shipping Thresholds: What Number Won't Lose You Money?

"Free shipping over NT$X" is probably the most gut-feel number in all of e-commerce. A competitor sets NT$990 so you set NT$990; customers complain about shipping fees so you drop it to NT$490; then the month-end margin report reveals free shipping ate more money than your ad spend. A free-shipping threshold isn't a marketing slogan — it's a math problem: your order-value distribution, your logistics cost structure, and how much more you want customers to add. Three variables, one threshold.

We've adjusted the free-shipping threshold on our own brand store several times, and every adjustment showed up directly in average order value and margin. This article lays the calculation out in the open.

Step One: See Your Logistics Cost Structure Clearly

List the true cost of every fulfillment method — not just the sticker shipping rate:

• Convenience-store pickup (7-Eleven/FamilyMart): the shipping fee plus size limits, and the two-way cost of unclaimed packages (no-pickup rate times round-trip shipping).
• Home delivery: tiered by volumetric size, surcharges for remote areas; fresh or fragile goods add packaging material costs.
• Cash on delivery: an extra collection fee, and a higher unclaimed-package risk.

Work out your average total logistics cost per order — that's your floor. Free shipping means moving that money from the customer's pocket to yours, so orders above the threshold must have enough margin to absorb it.

Step Two: Look at the Order-Value Distribution, Not the Average

This is the trap most people fall into: setting the threshold off "average order value." Averages get pulled up by large orders. What you should actually look at is the distribution — pull three months of order amounts and find the median and the range where most orders cluster.

The sweet spot for a threshold is usually 20 to 30% above the upper edge of the most common order-value range. For example, if most orders fall between NT$600 and NT$800, a threshold of NT$990 is "one small add-on away from free shipping"; NT$1,500 is so far away that customers give up on topping up — and often give up on the order. A threshold works as a reachable incentive; an unreachable one just feels like a penalty.

Set the threshold right and customers top up their carts willingly. Set it wrong and they don't buy more — they don't buy.

Step Three: Pave the Road to the Threshold

Setting a threshold without offering top-up options is digging your own hole. The cart page needs to do three things:

1. Show the gap: "Add NT$190 more for free shipping," updated in real time. That one line lifts order value more effectively than most banners.
2. Recommend top-up products: small items priced right in the common gap range (NT$100–300), with healthy margins and minimal package volume. If you don't have products like this, seriously consider developing one or two — their job isn't to make money, it's to push orders over the threshold.
3. Don't let top-up items break the package size: if topping up forces a bigger box and bumps shipping up a tier, you've defeated the purpose.

Incidentally, shipping fees are also a top cause of cart abandonment — revealing the fee for the first time at checkout is the classic abandonment-rate spike. State the shipping rules clearly on product pages and in the cart; this and cart abandonment recovery are two sides of the same coin.

Step Four: Validate and Adjust

After the threshold goes live, watch three numbers: whether the median order value drifts toward the threshold, the share of free-shipping orders, and per-order margin net of logistics costs. The healthy state is a cluster of orders "just over the threshold" — that means the incentive is working. If the free-shipping share is very high but order value hasn't moved, the threshold is too low and you're subsidizing orders that would have happened anyway. If almost nobody reaches it, it's too high and effectively decorative.

During major sale seasons you can temporarily lower the threshold as a campaign weapon, but do the math: lowering it means subsidizing shipping across the board, and like any discount it has a price. For the fuller margin-defense thinking, see designing for Double 11.

A final reminder: this whole calculation presumes your system can surface order-value distribution, unclaimed-package rates, and per-order logistics costs. If your current back office can't even pull those reports, the first thing to fix isn't the threshold — it's the data. That's why our e-commerce system service insists on structuring order and logistics-cost data for analysis. The era of pricing by gut feel should have ended long ago.