← Back to blog
Blog

Don't set your free-shipping threshold just above average order value — set it from how your orders are spread

Putting the free-shipping threshold a little above average order value works for some stores and quietly costs others margin, even when the average is identical. What decides it is how many orders sit just below the line. Here is how to count your orders by amount, pick a line, check how much shipping you can afford to absorb, and tell four weeks later whether it worked.

7 min read
  • Reading the numbers

This continues the third of the five numbers to check each week: average order value. That post listed revisiting your free-shipping threshold as one way to lift AOV. This one is the detail: how to decide the actual amount of the line.

The rule of thumb you hear most is “put it a little above your average order value.” It’s easy to apply, but two stores with exactly the same AOV can apply it and get opposite results — one sees bigger baskets, the other just gives away margin. What separates them isn’t the average. It’s how many orders are sitting just below the line.

Here are the things to check first, in order.

1. Same average, different place where orders pile up

Take two made-up stores. Each had 100 orders in the last four weeks, and each has an AOV of ¥4,000.

Order amount (excl. tax and shipping)Store AStore B
¥2,000–¥2,99950 orders15 orders
¥3,000–¥3,99915 orders30 orders
¥4,000–¥4,9995 orders40 orders
¥5,000 and up30 orders15 orders
Average¥4,000¥4,000

Both follow the rule of thumb and set free shipping at ¥5,000.

At Store A, only 5 orders sit just below the line, in the ¥4,000s. The pile is in the ¥2,000s, more than ¥2,000 away from the line — not a gap anyone closes by adding one item. Meanwhile, the 30 orders that were already above ¥5,000 now ship free without anyone changing what they buy. The store pays more shipping, and the shape of its orders barely moves.

At Store B, 40 orders sit in the ¥4,000s. They’re a few hundred to about ¥1,000 short. If there’s a good “one more item” to add, AOV goes up.

Both stores have the same ¥4,000 AOV. An average doesn’t tell you where the pile is. Under ③ in the five numbers post, AOV moved just because the mix of new and returning customers shifted. Same idea here: a single summary number hides the shape underneath.

Check: when you think about the threshold, look at your orders laid out by amount, not at AOV.

2. Count orders in ¥500 bands and look at the ¥1,000 just below the line

There’s one thing to do: count your recent orders by amount band.

  • Use four weeks to three months of orders. One week is too few to see a stable shape
  • Use ¥500 bands. If your prices span a wide range, ¥1,000 is fine
  • Count line-item totals excluding tax and shipping. Counting with shipping shifts everything near the line by exactly the shipping fee

Then, for a candidate line, look at what share of orders falls in the ¥1,000 just below it. Those are the customers one item away from free shipping. At the same time, look at the share of orders already above the line. Those customers won’t buy any differently; the store simply absorbs their shipping.

Slide the candidate line up and down in ¥500 steps and line up those two shares. You’re looking for an amount with a lot just below and not much above yet. That’s your first candidate.

One more thing: check that you actually sell something around ¥1,000 that can be added on its own. What gets a customer over the line isn’t the line — it’s the item they add. Without one, moving the line won’t get anyone across.

Check: for each candidate line, put “share in the ¥1,000 just below” next to “share already above,” and confirm you have an add-on item.

3. Use margin to cap how much shipping you can absorb

Why a free-shipping line works at all is clear enough. In Baymard’s survey of US online shoppers (as of November 2026), the top reason for abandoning a cart was extra costs such as shipping, tax and fees, at 40%, and not being able to see the total up front came in at 12%. Shipping is exactly the kind of cost that makes people stop.

But every order you ship free costs you margin. Before settling on a line, do this calculation once:

Shipping the store absorbs = shipping cost per order × number of orders above the line

That count includes not just the orders the line pulled over, but the ones that were already above it. At Store A in section 1, most of the shipping cost of a ¥5,000 line comes from the 30 orders that were already ¥5,000 or more.

If that total is bigger than the extra margin from the orders that crossed over (the margin on the added item), the line loses money. Lowering the line pulls more orders over, but it also puts more existing orders above it, so the cost grows fast. The thinner your margins, the less room you have to lower it.

Check: for a candidate line, compare the shipping you’d absorb with the margin the added items would bring. If the first is larger, don’t put the line there.

4. Four weeks later: AOV, order count, and the share near the line

When you move the line, write down the date, and four weeks later count the same bands again. Look at three things:

  • AOV — did it go up?
  • Order count — did it drop?
  • The share just below and just above the line — did orders that were just below move above?

AOV alone can mislead you. Say that after raising the line, AOV went from ¥4,000 to ¥4,400. Looks good. But if orders fell from 100 to 85, revenue went from ¥400,000 down to ¥374,000. Customers with small orders who couldn’t reach the line saw the shipping fee and didn’t buy. As in the traffic source post, a total moves when the mix inside it changes — and AOV rises when small orders simply drop out.

If the band just below got thinner and the band just above got thicker, the add-on worked. If not, you’re paying shipping for nothing, and the call is to move the line back. Keep the band counts from before the change so that you can.

Check: put AOV, order count and the share near the line side by side, before and four weeks after.

5. GA4’s standard reports don’t show how your orders are spread

Every decision above comes from the amount of each individual order. GA4’s standard reports have no screen that counts orders by amount band. You can get close in Explorations by listing revenue per transaction ID, but one problem remains.

Plenty of stores send the full amount paid, including tax and shipping, in GA4’s purchase event. In that case GA4’s revenue for each order is inflated by shipping and tax. And the thing you need to see for a free-shipping decision is precisely the few hundred yen just below the line. Shift that by the shipping fee and “just below the line” and “already above it” swap places.

So what’s missing for deciding the line is the per-order amount counted as line items excluding tax and shipping. Without it you can’t tell how many orders sit just below the line, and four weeks later you can’t tell whether to move it back. You need your cart’s order data.

On days when order data is connected, NextRise Analytics builds revenue and AOV from your cart’s orders, using line-item totals excluding tax and shipping. When orders tend to contain few items, it flags room for multi-item purchases on a card. Take a look at AOV and revenue by product on the Sales tab of the sample store.

If you’re not sure whether your GA4 receives the amount paid or the line-item total, Why your order count and conversion count never match explains how to check.

Related articles