Free shipping is a pricing decision

The three models: bake it in, pass it through, threshold

Shipping cost never disappears; it only moves. Every store is choosing, knowingly or not, one of three places to put it. Bake it in: raise item prices to absorb an average label cost and advertise free shipping. Buyers see one clean number, conversion improves, and the risk moves onto you — orders that ship cheaper than the average are extra margin, orders that ship dearer eat it. Pass it through: charge shipping at checkout, exactly or approximately. Every order carries its own freight and margins are protected, but the buyer meets a second number after emotionally committing to the first, which is where carts go to die. Threshold: charge below a line, free above it. The buyer is invited to solve the problem by buying more, and the bigger basket pays for the label.

None of the three is the correct one. Each is correct for a particular shape of catalogue: baked-in suits uniform, light, high-margin items whose label costs cluster tightly; pass-through suits heavy, bulky, or wildly variable parcels where no average is honest; thresholds suit catalogues where a plausible second item exists. The decision inputs are your average order value, your average label cost, and your margin — three numbers this page assumes you know, and the last of which you only know precisely when every label is bought at one all-in price.

The margin math, worked on one real order

Abstractions hide the trade, so here is one order run through all three models. The shape to notice before the table: the buyer's total barely moves across the first two columns, yet the checkout experience is completely different — one store looks generous and one looks stingy, on identical economics. The threshold column is the interesting one: the seller absorbs a slightly bigger label and more cost of goods, and still comes out ahead on margin because the basket grew. That is the entire threshold thesis in one row of arithmetic.

Worked example — one illustrative order, priced mid-2026. The item, goods cost, and margins are invented for the arithmetic; the label figures reflect a typical light domestic parcel at commercial pricing, verified against the carriers' published commercial rate charts on 2026-08-15. Label costs move with annual rate changes and with your parcels' weight and zones, so rerun this table with your own ledger numbers before acting on it — the method is the asset, not these dollars. Re-checked at the carriers' annual rate announcements by the named marketing owner recorded with this page.

Setting a threshold from your average order value

The threshold formula is short enough to do on a receipt. Call your current average order value AOV, your average label cost L, and your contribution margin rate m — the share of each product dollar that is profit before shipping. Place the threshold T a step above AOV, and the free-shipping deal breaks even when the extra margin from the bigger basket covers the label you are now absorbing: it pays for itself whenever m × (T − AOV) ≥ L. Rearranged, the lowest defensible threshold is T = AOV + L ÷ m. Anything below that line is a discount wearing a shipping costume; anything far above it stops changing behaviour because buyers cannot see a realistic path to reaching it.

Two refinements make the formula honest. First, use the label cost of the basket that actually crosses the threshold — an extra item adds weight, and a heavier parcel can cross a weight tier. Second, treat the formula as a floor, then round to a number that reads well on a banner. Rerun it quarterly: AOV drifts, and L drops the moment you start rate-shopping every parcel, which lowers the threshold you can afford and makes the offer stronger at the same margin. The mechanics of pushing L down are the cheap shipping labels page's whole subject, and keeping fixed tooling costs out of L entirely is what the no-monthly-minimum pricing model is for. One more input if you sell on a marketplace: several of them badge or boost free-shipping listings in search, so the model doubles as a visibility decision — the Etsy-specific version of that trade lives on the shipping labels for Etsy page.

When buyer location wrecks the model, and what to do about it

Flat and free models price on averages, and averages have a failure mode: the buyer who orders from the far end of the country. Distance bands move the same parcel's label cost substantially between nearby and cross-country destinations, and destination fees stack on top for residential and harder-to-reach addresses — the shipping surcharges page owns those triggers, and the assembled anatomy of how a rate is built lives on the rate calculation page. A free-shipping banner priced on your best-case zone is a coupon your most distant buyers redeem every time.

The structural fixes are three. Price the model on the zone-weighted average of what you actually shipped last quarter, not on the label cost you remember. Fence the outliers: keep calculated shipping for the few heavy or oversized items where variance is genuinely unpriceable, and let the ordinary catalogue stay flat or free. And shrink the variance itself by quoting every parcel across carriers — the carrier that wins a nearby zone is often not the one that wins a distant one, and picking the winner per order narrows the spread between your cheapest and dearest deliveries; the ground-tier crossover detail is the cheapest ground services page's. A narrower spread makes every model on this page safer, which is the quiet way the label tool and the pricing strategy turn out to be the same decision.