Ecommerce shipping strategy: the free shipping math

Shipping is the line in the checkout where people change their minds. Not the price of the product — they agreed to that on the product page — the $6.95 that appears afterward and makes the total feel like a different decision.

So the standard advice is "offer free shipping". That advice is right about psychology and silent about arithmetic, which is a problem, because free shipping is not a shipping decision. It is a price cut, and a large one.

This post is the arithmetic. Every figure in it is illustrative — invented so the structure is visible — and the point is that you redo each calculation with your own. Substitute your average order value, your real carrier costs and your own processing rate wherever they appear.

The one number everything turns on

Not revenue, not gross margin. Contribution per order: what is left after the cost of the goods, the cost of shipping them, and the payment fee.

The illustrative store: average order value $38, cost of goods 45% of the product price (so a 55% gross margin), average real cost to ship an order $7.50, shipping charged to the customer $6.95, and payment processing of 2.9% plus 30 cents — an illustrative assumption, not a quote from any provider, so put your own in.

That $18.75 is the number to protect. Everything below is measured against it.

What switching to free shipping actually does

Same order, same $38 price, shipping now free.

Contribution falls from $18.75 to $12.00. That is $6.75 gone from every order, or 36% of what you were keeping.

Now the question that matters: how many more orders do you need for the trade to be worth it? Divide the old contribution by the new one — 18.75 / 12.00 = 1.56.

You need 56% more orders just to end up where you started. Not 5% more. Not 10%. Free shipping has to grow your order count by more than half before it earns you a cent.

That is the sentence the standard advice leaves out. It is not impossible — a visible free shipping message genuinely does move conversion — but 56% is a very large effect to assume, and assuming it wrongly means you have cut your own margin by a third in exchange for nothing.

Free shipping is a price change, so price for it

There is a third option most stores skip: keep shipping free and raise the product price to cover it.

Exactly the same $18.75 as charging shipping separately. Which tells you something useful: the real question was never free versus paid shipping. It is whether $44.95 with free shipping sells better than $38 plus $6.95 — and those two carts cost the customer the identical amount.

Usually the bundled price does sell better, because it removes the surprise at the last step and because people compare product prices long before they compare totals. Sometimes it sells worse, because $38 clears a mental threshold that $44.95 does not. That depends on your category, and you can only find out by trying.

The important part is that this version costs you nothing to try. Which makes it the first thing to test, not the last.

What a flat rate really costs

Flat rate looks like a compromise and behaves like a quiet subsidy.

Say you charge a flat $5.95 nationwide. Your actual costs are not flat: nearby orders cost about $5.50, distant or heavier ones about $11.00, and roughly 30% of orders fall into the expensive group. The blended cost is (0.70 x 5.50) + (0.30 x 11.00) = 7.15.

You collect $5.95 and spend $7.15, so you are paying $1.20 an order for the simplicity. That may be worth it — one number on the page is easier to trust than a calculator. But it is a cost, and it grows every time a distant customer finds you.

Set a flat rate from the blended cost, not the cheapest zone. A flat $7.50 breaks even on this store. A flat $5.95 is a discount you did not decide to give.

Where to set a free shipping threshold

A threshold is the only version of free shipping that can reliably pay for itself, and it does so only if the number is right.

Two things happen when you set one. Orders that get bigger in order to qualify earn you extra margin on the added items. Orders that were already above the number get free shipping they would happily have paid for.

So each uplifted order gains $2.10 and each already-qualifying order loses $6.75. Divide one by the other — 6.75 / 2.10 = 3.2.

You need a little over three uplifted orders for every order that would have qualified anyway. That single ratio is the whole threshold decision, and it explains the rule everyone repeats without its reasoning: set the threshold above your average order value. If your AOV is $38 and the threshold is $35, almost every order qualifies by accident, nothing uplifts, and you have simply given away $6.75 a time.

A workable starting point is 1.3 to 1.5 times your average order value, with one condition attached: the gap must be closeable by adding one ordinary item. A $55 threshold in a store where everything costs either $12 or $90 uplifts nothing, because there is nothing to add that lands in the gap.

Three places free shipping is simply wrong

Heavy or bulky goods. Replace the $7.50 with $18.00 on that $38 order and the free-shipping version leaves $1.50 of contribution. No volume lift repairs a structure like that. Charge for freight and say why in one sentence.

Low margin. At a 30% gross margin, a $38 order leaves $11.40 of product margin, which does not cover $7.50 of shipping plus a payment fee. Free shipping on thin margins is not aggressive pricing, it is loss-making pricing.

Across borders. International costs vary by multiples rather than percentages, and duties are a separate shock you cannot absorb on the customer's behalf. Charge calculated rates and be explicit about who pays duty.

What this math cannot tell you

It cannot tell you how much free shipping lifts your own conversion rate. That is the one input deciding whether the trade is good, and it is specific to your store, your category and your price points. Anyone handing you a universal percentage is handing you an average of stores that are not yours.

It also cannot be measured honestly at low volume. At forty orders a month, the gap between a 2.0% and a 2.6% conversion rate is a couple of orders — well inside ordinary weekly noise. If that is you, do not run a shipping test at all. Adopt the margin-neutral option and spend the week on traffic instead; getting your first customers is the better use of it.

And it says nothing about returns. If you pay return freight as well, a refunded order costs you shipping twice and earns no margin at all. That decision belongs in your return policy, made deliberately.

A sane order to do this in

1. Work out your real contribution per order, using the blended shipping cost rather than the cheapest zone.
2. Move to free shipping with the price raised to cover it. Margin-neutral, so it costs nothing to try.
3. If your average order is smaller than you want, add a threshold at 1.3–1.5 times AOV and check it against the ratio above.
4. Charge calculated rates for anything heavy, bulky or crossing a border, and explain it at checkout in one line.
5. Re-run all of it when carrier prices change, which they do every year.

A whole ecommerce shipping strategy fits in those five lines. The reason it is worth writing down is that every line is a number you can check, and the advice it replaces was a feeling.

mAutomate covers the operational side of this — orders, inventory, sales and analytics in one dashboard — and the AI drafts changes and recommendations that wait for your approval before anything goes live. The arithmetic above stays yours, because only you know what a lost order is worth to you. The plans page lists what each tier includes, and every plan starts with a 14-day free trial, no card required.