Free Shipping Thresholds Are Also a Zone Problem

Free Shipping Thresholds Are Also a Zone Problem

Most threshold math treats shipping cost as one number. It isn't. A near-zone order and a far-zone order can differ by 2× or more in carrier cost — which means a single national threshold is simultaneously too generous and too stingy, depending on the zip code.

The windfall you can't see

Take a typical mix: 60% near-zone at ~$6.50, 40% far-zone at ~$12. Blended average: $8.70.

Now a $75 threshold where 40% of orders already clear it. Those always-qualifiers get free shipping without changing their behavior — and if 40% of them are far-zone, that's 160 orders × $12 = $1,920 of windfall subsidy per 1,000 orders, before a single shopper stretches a basket.

Standard threshold analysis (top-ups minus subsidies) usually assumes one carrier cost. With zone mix, the subsidy side is dominated by far-zone always-qualifiers — and the "profitable threshold" quietly isn't.

Why one threshold can't win

The optimum sits where the margin from one more dollar of top-up capture equals the subsidy lost on one more qualifying order. But those two things have different zone profiles:

• Top-ups skew near-zone — lower baskets, shorter distances, impulse adds.
• Always-qualifiers skew everywhere, including far zones where the subsidy is double.

A single number splits the difference and gets both wrong: near-zone shoppers clear it without trying (windfall), while far-zone shoppers face a number that, paired with their shipping cost, rarely converts.

What to do instead

1. Compute the threshold separately by zone cluster. Near and far can each have a number that pays for itself — a lower threshold near, a higher one (or a handling offset) far.
2. Watch always-qualify drift. As AOV inflates, more orders clear the threshold by default. Alert when the always-qualify share rises five points — your windfall is growing silently.
3. Count split carts. Some shoppers split orders to get two free-ship parcels. Measure per customer-week, not per order, and add anti-split logic.
4. Net stretched add-ons of refunds. Stretch items return at higher rates; reduce the "top-up gain" by their refund rate before declaring victory.
5. Re-estimate quarterly. Zone mix, carrier rates, and AOV all drift. A threshold set once decays.

The takeaway

A free-shipping threshold isn't one decision. It's one decision per zone mix — and the version that ignores zones subsidizes the far and taxes the near.

Set the threshold where the math works for the zone that pays for it, watch the always-qualifiers, and never let a national number decide a local cost.