A free-shipping threshold changes behavior for two different groups of shoppers. A good threshold is one where the gain from the first group outweighs the cost of the second.
Start with your order distribution.
Average order value hides the shape you need. Export recent orders and look at how many fall in each price band: under $40, $40–$60, $60–$80, and so on. A threshold works by nudging people who are just below it, so you need to know how many people are just below each candidate number.
Model the shoppers who add something.
Here is an illustrative example, not a benchmark. Today you charge $7 for shipping and it costs you $9. You test free shipping on orders of $75 or more.
A shopper with a $62 cart adds a $15 item to qualify. If that item has a 60% gross margin, the add-on contributes $9. You give up the $7 shipping charge you used to collect. Net effect on that order: about +$2, plus a customer who got something they wanted.
If the add-on has a low margin, or the extra item makes the parcel more expensive to ship, the same order can come out negative. Use the margin of the items people actually add, not your store average.
Model the shoppers already above it.
A customer who was going to spend $90 anyway now pays no shipping. In this example you lose the $7 charge on every one of those orders and gain nothing in return. If a large share of your orders already sit above the threshold, this group can cost more than the first group earns.
Put the two together.
- Orders that move up to qualify × the net gain on each
- minus orders already above the line × the shipping revenue you give up
- plus any change in conversion rate you can actually measure
Run that for two or three candidate thresholds. Often the best choice sits a little above the most common order value, close enough that one sensible add-on reaches it.
Make qualifying easy.
A threshold only works if people know about it and can reach it. Show the remaining amount in the cart, suggest relevant, in-stock add-ons near the right price, and state the shipping policy plainly on product pages so it is not a surprise at checkout.
Treat the threshold as a test with a start date, a review date, and the numbers above written down in advance. Then compare contribution, not only revenue.
Want a second pair of eyes?
Use the free worksheet on your own store, or book a 15-minute call to talk through your situation.