How to Set a Free Shipping Threshold From Your Own Order Data

The usual rule for a free shipping threshold is a percentage above average order value. Your order value histogram can do better: it shows whether shoppers already top up to reach your threshold, how many orders sit just below it, and where contribution per visitor peaks.

Most advice on choosing a free shipping threshold comes down to one rule: take your average order value and add a margin. Eightx, a finance firm for ecommerce brands, sums up the usual version as setting the threshold 15 to 25 percent above current AOV and then checking it against shipping cost and gross margin. It is a reasonable starting point. It is also a guess, because an average says nothing about where your orders actually sit.

The threshold is worth getting right. Shipping cost is the most cited reason for leaving a checkout: in Baymard Institute’s February 2024 survey, 48 percent of US adults who abandoned a cart did so because extra costs such as shipping, tax and fees were too high. This guide uses the orders you already have to answer three questions: does your current threshold change behaviour, how many orders are close enough to be nudged, and which threshold earns the most per visitor rather than per order.

Why average order value is the wrong input

A threshold only acts on orders in a narrow band below it. A shopper with $20 in the cart will not add $40 to save $8 of shipping, and a shopper at $140 is already over the line. The average hides how many orders fall in the band that matters, and it is pulled around by a minority of large orders. Device mix is one example: Metorik’s 2026 WooCommerce benchmark puts desktop AOV at $167 against $71 on mobile, with mobile placing 72 percent of orders. A threshold set from the blended average can sit far above where most mobile baskets end.

Be careful with the obvious comparison, too. The same Metorik data shows that 74 percent of WooCommerce orders included free shipping in 2025, and free-shipping orders averaged $123 against $80 for orders that paid for shipping. That gap does not prove free shipping raises order value. Orders above a threshold are larger by definition, so the comparison would show a gap even if nobody changed their basket. You need a method that separates selection from behaviour.

Step 1: Plot the order value histogram

Group a few months of paid orders into $5 bins by the order value your shipping rule uses, and plot the counts. If your threshold is working, you will see a pile of orders just above it and a shortage just below it. Economists call this pattern bunching, and there is a well-developed method for measuring it, summarised in Henrik Kleven’s review Bunching in the Annual Review of Economics (2016).

Order value histogram with bunching at a 60 dollar threshold Bars rise from low order values to a peak around 35 dollars, then decline. Just below the 60 dollar threshold the bars dip under a smooth dashed curve, and just above it the bars jump well above the curve before falling back. Threshold: $60 Missing orders Excess orders (shoppers who topped up) Smooth counterfactual $15$35$55$60$75$90 Order value (after discounts, before shipping and tax), $5 bins
Illustrative. The dark bars just above the threshold rise well above the smooth curve fitted to the rest of the distribution. That excess is the evidence that the threshold changes baskets.

Pull the counts with a query along these lines on WooCommerce:

-- Paid orders grouped into $5 bins of net order value
-- net_total excludes shipping and tax and is after coupon discounts.
-- If your free shipping rule is applied BEFORE coupons, add the
-- coupon discount back (see wc_order_coupon_lookup).
SELECT FLOOR(net_total / 5) * 5 AS bin_start,
       COUNT(*)                 AS orders
FROM wp_wc_order_stats
WHERE parent_id = 0
  AND status IN ('wc-processing', 'wc-completed')
  AND date_created >= '2026-06-01'
  AND date_created <  '2026-11-01'
GROUP BY bin_start
ORDER BY bin_start;

The comment matters. WooCommerce’s free shipping method has a setting to apply the minimum order rule before coupon discount. With the box checked, a $30 order with a $10 coupon counts as $30 for the threshold. Bin your orders on the same definition the rule uses, or the bunching will appear in the wrong place and look weaker than it is.

Step 2: Measure the excess

The bunching estimate is simple to compute in a spreadsheet:

  1. Pick an exclusion window around the threshold, for example from $10 below it to $15 above it.
  2. Fit a smooth curve (a polynomial of degree 4 to 7 works) to the bin counts outside that window.
  3. Use the curve to predict counts inside the window. That prediction is what the distribution would look like without a threshold.
  4. Excess mass is the sum of actual minus predicted counts for bins above the threshold. Missing mass is the same sum for bins below it. In a clean case the two are roughly equal, because the orders that bunch above came from just below.

Divide the excess by the number of orders the curve predicts in the band just below the threshold. That ratio is your top-up rate: the share of near-miss shoppers who added items to qualify. Survey figures for this are around 58 percent, but those come from asking people, not from watching their carts. Your own top-up rate is the number to use in the model below, and it is often much lower.

What you seeWhat it suggestsNext step
Clear pile just above the threshold, gap just belowThe threshold changes basketsMeasure the top-up rate and test a higher threshold
No visible bunchingShoppers do not notice the threshold or cannot reach itCheck it is shown in the cart, then test a lower threshold
Bunching at a value other than your thresholdA popular product price or bundle, or the rule uses a different order value than you thinkCheck the before or after coupon setting
Very few orders in the $10 below the thresholdThe threshold sits where there is little to nudgeMove it nearer a dense part of the distribution

Step 3: Model contribution per visitor, not per order

For each candidate threshold, every order falls into one of three groups: above the threshold (you pay the parcel), within reach below it (some share tops up, the rest pay shipping), and far below it (they pay shipping as before). Contribution per order is margin on the basket, minus the shipping you absorb, plus any shipping revenue you collect.

Eightx published a worked version of this model for a hypothetical brand with a $60 AOV, a 60 percent contribution margin and a $14 parcel. Their table shows contribution per order rising from $22 with unconditional free shipping to a peak of about $37 at a $70 threshold, then easing off as fewer orders qualify. The author states that the curve is illustrative and that real thresholds should come from a brand’s own actuals.

Net contribution per order by free shipping threshold, Eightx worked example Line chart. Net contribution per order is 22 dollars with no threshold, 28 at 50, 32 at 60, peaks at 37 at 70, then 36 at 75, 34 at 90 and 33 at 110. $20$25$30$35 $22$28$32$37$36$34$33 None$50$60$70$75$90$110 Free shipping threshold (Eightx example: $60 AOV, 60% margin, $14 parcel)
Data from Eightx’s illustrative model. Your curve will have a different shape and peak; the method is what carries over.

That model works per order. Add one more term before you trust it: conversion. A higher threshold can lift contribution per order while losing shoppers who never reach checkout, so the figure to maximise is contribution per visitor (conversion rate multiplied by contribution per order). The inputs you need, and where to find them:

InputWhere it comes fromCommon mistake
Order value distributionOrder table, binned as in Step 1Using AOV instead of the distribution
Top-up rate and reachBunching estimate from Step 2Using a survey figure
Contribution margin by basketProduct costs, payment fees, packagingUsing gross margin and forgetting fees
Parcel cost per orderCarrier invoices by weight and zoneUsing the rate charged to customers instead of the rate paid
Shipping revenue below the thresholdOrder table (shipping total)Ignoring it; Metorik’s 2026 data puts the average paid shipping charge at $9.27
Checkout conversion at each thresholdA test (Step 4)Assuming it stays constant

Step 4: Test the change

A model narrows the options. A test settles them. Two designs fit most stores:

  • By shipping zone. WooCommerce sets free shipping per zone, so you can raise the threshold in some regions and keep it in comparable ones. Compare conversion rate, order value and contribution per visitor across the two groups over the same weeks. This avoids most seasonality problems.
  • Before and after. Change the threshold everywhere and compare four or more weeks on each side, matching weekdays. Simpler, but any promotion or seasonal shift in the window will be read as the threshold’s effect.

Decide the sample size and the length of the test before starting, and do not stop early because the first week looks good. QuantumLayers covers why in A/B Testing Statistics. After the change, rerun the histogram. The bunching should move with the threshold. If it does not, shoppers are not seeing the new number.

What to monitor after you change it

MetricWhy it matters
Contribution per visitorThe number the threshold should maximise
Checkout conversion rateCatches shoppers lost to a threshold that is too high
Share of orders within $15 below the thresholdShows how many orders are still being nudged
Excess mass above the thresholdConfirms the threshold still changes behaviour
Absorbed shipping cost per orderCarrier rates change; the model’s parcel cost goes stale
Return rate on topped-up ordersFiller items added to qualify are sometimes sent back

Sources