Setting a Free Shipping Threshold: The AOV Math Most Stores Skip

Set the threshold above your median order value, close enough that a typical basket can reach it with one added item, and low enough that the shipping cost you absorb stays covered by the margin on the enlarged basket. In practice that lands most stores 15 to 30 percent above their current median order. A threshold at your average order value changes nothing, because half your customers already clear it. A threshold double the median is a poster, not an incentive; shoppers do the same gap math you should have done and decline.

Thresholds get picked in odd ways: the founder's round number, the biggest competitor's number, the number from a theme's default progress bar screenshot. Then the store wonders why free shipping uptake is high and profit is down, or why nobody stretches and the threshold might as well not exist. Both failures are arithmetic that was never run.

Your order history already contains the answer. The distribution of basket values tells you where a threshold bites, and your carrier bill tells you what each redemption costs. The only judgment call left is how much stretch a customer will make, and even that is observable in your current data around any existing minimums.

Read the distribution, not the average

Export twelve months of order values and look at the shape. Averages lie because a few large orders drag them up; the median and the cluster points are what matter. If orders bunch at 38 and 64 dollars, a threshold at 55 sits in the empty valley, closeable from the 38 cluster with one mid priced item and already cleared by the second cluster. That is a threshold with a job.

Now price the redemption honestly. Your absorbed cost is the shipping you would have charged, minus nothing; use your real average carrier cost per parcel, not the discounted rate in the pitch deck. Multiply by the share of orders you expect to clear the line after it moves behavior, including the orders that already cleared it for free on day one. That last group is the silent cost most threshold launches forget to budget.

The stretch gap is smaller than you hope

Shoppers will add an item to close a gap that feels like a rounding error relative to the basket, roughly a fifth to a quarter of the order value in most observed behavior, and they balk beyond it. A 40 dollar basket will stretch to 50. It will not stretch to 75 because a banner asked nicely. If your threshold needs a bigger stretch than that from the median basket, lower the threshold or accept that it is decoration.

The add on item matters too. A threshold strategy quietly depends on having a 12 to 20 dollar product worth adding: the consumable, the accessory, the second unit. Stores whose catalog jumps from 40 dollar items to 90 dollar items cannot run a close threshold at all, and should fix the assortment gap before touching the banner.

Test it as a price change, because it is one

Announcing free shipping over a threshold changes your effective prices. Run it like a pricing decision: one segment or one period, watch conversion, average order value, and contribution margin per order together, never conversion alone. A threshold that lifts orders 6 percent and cuts margin per order 9 percent is a promotion you cannot afford wearing an operational costume.

Seasonality will move your numbers under you. The threshold set against November baskets, fat with gifts, will sit out of reach of your thinner March baskets. Set it against the full year distribution, and resist re-tuning it monthly; customers learn thresholds, and a moving line teaches them to wait rather than stretch.

Frequently asked questions

Should the threshold be a round number?

Yes, for memory and banner clarity, but choose the round number nearest your computed target rather than starting from the round number. Fifty or seventy five are fine destinations for the math; they are bad starting points.

Do I count the threshold before or after discounts?

Decide once and say so on the banner and the cart. Most stores count the merchandise subtotal after discounts, before shipping and tax. Whatever you choose, a customer whose code drops them below the line mid checkout needs a clear message, not a vanished benefit at the payment step.

Is free shipping over a threshold better than flat rate shipping?

For basket building, usually yes; the threshold converts shipping from a cost to be minimized into a goal to be reached. Flat low rate shipping converts better for single item stores where no plausible threshold sits within stretch distance of the median basket.

How long before I can judge a new threshold?

Give it at least several hundred orders per segment and a full weekly cycle, longer if your store has strong weekday seasonality. Judging on the launch weekend measures novelty and your email announcing it, not the threshold.