Every guide to free shipping thresholds gives you the same two instructions: use the break-even formula, then set the number 15 to 25% above your average order value. Both are reasonable. Neither tells you what the threshold costs, and the largest cost in the model is the one nobody puts in it. With carrier rates rising on 4 January, this is a good month to put the whole thing on a spreadsheet rather than a rule of thumb.
Start with break-even threshold = average order value + (shipping cost ÷ gross margin). At a $62 AOV, 55% margin and $8.50 shipping, that is $77.45, so $79 is marginally profitable and $75 is marginally subsidised. Then test it against your order-value distribution rather than your average, because the real cost of a threshold is the shipping you absorb on orders that would have qualified anyway. On the numbers below, a $75 threshold needs roughly a 7% lift in order volume just to break even.
Why this is on the clock
FedEx announced its 2027 general rate increase in September 2026: an average of 5.9%, effective 4 January 2027, the fourth consecutive year at that headline figure. UPS has historically followed within a few weeks and had not published its 2027 increase as of early October.
The headline is the least useful number in the announcement, because the surcharges that direct-to-consumer parcels actually attract are rising faster than the base rate:
| FedEx 2027 increase | Rise | Who it hits |
|---|---|---|
| Average rate increase (headline) | 5.9% | Everyone, as an average across a mix most shippers do not have |
| Additional handling | 6.8%–7.6% | Anything awkward: long, dense, irregular packaging |
| Oversize | 5.5%–5.9% in zones 2–4, 7.6%–7.8% in zones 5 and above | Bulky goods shipped long distances |
| Extended residential delivery area surcharge | 9.1% | Rural and outer-suburban consumers |
| Residential pickup | 9.2% | Brands shipping from home or a small unit |
| Five-day regularly scheduled pickup | 9.7% | Any store with a daily collection |
| Declared value | 9.1% | Higher-value goods |
A small DTC brand shipping residential parcels on a scheduled daily pickup, some of them to extended delivery areas, is not facing 5.9%. It is facing something closer to 7%, and the only way to know is to apply each increase to your own service and surcharge mix. That matters here because a threshold is priced in your shipping cost, and a threshold set against 2026 costs starts quietly losing money in January.
Extra costs are also the thing shoppers abandon over most. Baymard's research, based on a survey of 4,384 US adults, puts 48% of abandonments down to extra costs such as shipping, tax and fees being too high, against an average cart abandonment rate of 70.22% across 50 studies. So the threshold is simultaneously your biggest margin lever at checkout and your biggest abandonment trigger, which is exactly why guessing at it is expensive.
The formula, and what it leaves out
The standard calculation is sound. You are looking for the cart value at which the extra margin on what a customer adds covers the shipping you give away:
Threshold = AOV + (shipping cost ÷ gross margin)
At $62 AOV, 55% gross margin and $8.50 blended shipping: 62 + (8.50 ÷ 0.55) = 62 + 15.45 = $77.45.
And the separate floor test, for whether an order is contribution-positive at all once you absorb delivery: shipping ÷ margin = 8.50 ÷ 0.55 = $15.45 of merchandise.
What the formula assumes is that the threshold's only cost is the shipping on the orders it persuades. That is not its only cost. It is not even its largest one. Every order that was already above the line now ships free too, and those customers were going to buy anyway.
The common "1.3 to 1.5 times AOV" rule has a second blind spot: two stores with the same $62 average order value can have completely different distributions behind it. One with most orders clustered between $55 and $70 has a large persuadable population just below a $75 line. One with a barbell of $25 orders and $140 orders has almost nobody in reach, and a threshold that mostly subsidises the top of the barbell. The average is identical. The right threshold is not.
The cart-value band model
This is the part worth building once. Pull your last 90 days of orders, bucket them by value, and work in orders per thousand so the arithmetic stays legible. A worked example, at 55% margin and $8.50 blended shipping:
| Order value band | Share of orders | Role at a $75 threshold |
|---|---|---|
| Under $40 | 18% | Out of reach. A $35 gap is not an upsell, it is a different order |
| $40–$54 | 22% | Out of reach for most |
| $55–$64 | 20% | Partly persuadable at the top of the band |
| $65–$74 | 15% | Persuadable. Within $10 of the line, which is where add-on behaviour concentrates |
| $75–$99 | 16% | Leakage. Qualifies automatically; you now absorb shipping you were charging for |
| $100+ | 9% | Leakage. Same, on your best orders |
Now price it, per 1,000 orders, for a store moving from charging shipping to a $75 threshold:
| Line | Working | Value |
|---|---|---|
| Orders already above $75 | 25% of 1,000 | 250 |
| Subsidy on orders that qualified anyway | 250 × $8.50 | −$2,125 |
| Persuadable orders ($60–$74.99) | 25% of 1,000 | 250 |
| Orders persuaded to cross, at a 25% crossing rate | 250 × 25% | 63 |
| Extra margin from those crossings, at $10 added each | 63 × $10 × 0.55 | +$347 |
| Subsidy on the persuaded orders | 63 × $8.50 | −$536 |
| Net, before any new orders | −$2,313 |
Read the second line first. The subsidy on orders that would have happened anyway is $2,125, roughly six times the margin gained from everyone you persuaded to add something. Free shipping thresholds are almost always sold on the upsell effect. The upsell effect is the small number in the model.
The number that actually decides it
A threshold does one more thing the table above has not counted: it wins orders you would otherwise have lost, because extra costs are the leading abandonment reason. So the honest question is not "does the upsell cover the subsidy", which it rarely does. It is how much extra order volume the threshold has to produce to pay for itself.
Required additional orders = net shortfall ÷ contribution per additional qualifying order
Contribution per additional order at $75 = (0.55 × 75) − 8.50 = $32.75.
Required additional orders = $2,313 ÷ $32.75 = 71 per 1,000, a 7.1% lift in order volume.
That is a test you can actually apply. Reported ranges put abandonment reduction from a threshold at 15 to 25%, which if it held on your store would clear a 7% hurdle comfortably. But those ranges come from stores that mostly had no free shipping offer at all and often sell in categories where shipping is a larger share of the order. If your checkout already converts well and your traffic is mostly returning customers, a 7% volume lift from a shipping policy change is a strong claim, not a safe assumption.
Run the same model at a higher threshold and something counterintuitive appears:
| Line | $75 threshold | $85 threshold |
|---|---|---|
| Orders qualifying anyway | 250 | 170 |
| Leakage subsidy | −$2,125 | −$1,445 |
| Orders persuaded to cross | 63 | 33 |
| Extra margin from crossings | +$347 | +$200 |
| Subsidy on persuaded orders | −$536 | −$281 |
| Net shortfall | −$2,313 | −$1,526 |
| Contribution per additional order | $32.75 | $38.25 |
| Break-even volume lift | 7.1% | 4.0% |
The higher threshold persuades half as many people and has a break-even hurdle almost half as high, because leakage shrinks faster than persuasion does. This is the opposite of what the "set it just above your AOV" advice implies, and it is the single most useful output of building the model on your own data. A threshold set close to your average order value maximises the number of customers who feel rewarded, and maximises the number you were paying to reward for nothing.
Subsidy leakage = (orders that would have exceeded the threshold anyway) × shipping cost absorbed. Measure the first term from your pre-threshold distribution, not your current one, because the threshold itself has pushed some orders above the line. Report it monthly in currency, not as a percentage: the percentage rises as you raise the threshold even while the money falls, which is how a well-meant dashboard can tell you the opposite of the truth.
Which policy, when
A threshold is one of four reasonable answers, and for some cost structures it is the wrong one.
| Policy | Choose it when | Avoid it when |
|---|---|---|
| Threshold | A meaningful share of orders sits within $5–$15 of a credible line, and your margin covers shipping on an added item | Your distribution is a barbell with few carts in reach, or margin is thin enough that the added item does not pay for delivery |
| Free shipping on everything | Margin on your smallest realistic order still clears the floor test, and simplicity is worth more than the subsidy | You sell low-value items, or shipping cost varies widely by weight and zone |
| Flat rate | Shipping costs vary too much to absorb safely, but you want the certainty a quoted price gives shoppers | Competitors in your category have normalised free delivery and the flat fee reads as a penalty |
| Real-time carrier rates | Heavy, bulky or international orders where absorbing cost is impossible | Your average basket is small; a calculated fee appearing late is the classic extra-cost abandonment |
Whichever you choose, show it early. The abandonment data is about unexpected costs, not costs as such, which is also the direction fee-disclosure regulation has been moving, as we covered in all-in pricing laws and checkout. A flat $7 shipping fee stated on the product page damages conversion far less than a $7 fee discovered at step three.
The January repricing procedure
- Pull your true blended shipping cost per order, including every surcharge, for the last 90 days. Not your carrier's base rate card, and not what your shipping app estimates. The number you need is total carriage divided by orders shipped.
- Apply the increases to your own mix, line by line, rather than applying 5.9% to everything. Residential, DAS and scheduled-pickup items rise 9% or more, so most DTC shippers land above the headline.
- Recompute the break-even threshold with the new cost. On our example, shipping moving from $8.50 to $9.10 lifts break-even from $77.45 to $78.55, and the break-even volume lift at a $75 threshold from 7.1% to 7.8%. Small movements, compounding annually, which is how a threshold set in 2024 ends up subsidising a business in 2027.
- Re-check the bands, because your distribution has moved too. If your current threshold has pulled a cluster of orders just above the line, the next threshold should be set against the pre-threshold shape, or you will chase your own tail upwards.
- Change the number in January, and only then. A threshold change moves conversion rate and average order value simultaneously, so it is unreadable against peak-season traffic and the downside risk is at its worst. This is exactly the kind of change that belongs in the medium-risk tier of a tiered peak-season freeze: ship it in September or wait until the second week of January.
One further thing worth testing separately rather than bundling in: progress messaging. "Add $9 for free shipping" is a different intervention from the threshold itself, it is cheap to implement, and it affects the crossing rate directly, which is the one term in the model you can improve without giving away more margin. Treat it as its own test, sequenced after the threshold is right, in the same way we would sequence any of the items in the cart abandonment recovery stack.
Frequently asked questions
How do you calculate a free shipping threshold?
Use break-even threshold = average order value + (shipping cost divided by gross margin). At a $62 average order value, 55% gross margin and $8.50 blended shipping cost, that is 62 + (8.50 / 0.55) = $77.45, so a threshold of $79 is marginally profitable and $75 is marginally subsidised. The formula finds the cart value at which the extra margin on the items a customer adds covers the shipping you absorb.
Is the rule to set a free shipping threshold 1.3 to 1.5 times your AOV correct?
It is a starting point that ignores two things that decide the outcome. It ignores your gross margin, which determines whether an added item actually pays for the shipping, and it ignores the shape of your order-value distribution, which determines how many carts sit close enough to the line to be persuaded. Two stores with identical average order values and different distributions should set different thresholds.
What does a free shipping threshold actually cost?
Mostly it costs you the shipping you now absorb on orders that would have qualified anyway, which is the largest single line in the model and the one usually left out. On 1,000 orders with 25% already above a $75 threshold and an $8.50 shipping cost, that subsidy is $2,125 before a single customer has been persuaded to add anything. The persuaded orders have to cover it, and in most cost structures they do not on their own.
What order volume lift does a free shipping threshold need to break even?
Work it out from your own numbers: the shipping you absorb on all qualifying orders, less the extra margin from the carts you persuade to cross the line, divided by the contribution on one additional qualifying order. On a worked example with a $62 average order value, 55% margin, $8.50 shipping and 25% of orders already above a $75 threshold, the shortfall is about $2,313 per 1,000 orders and each additional qualifying order contributes $32.75, so the threshold needs roughly a 7% lift in order volume to break even.
Are shipping costs going up in 2027?
Yes. FedEx announced its 2027 general rate increase in September 2026 at an average of 5.9%, effective 4 January 2027, the fourth consecutive year at that headline level. The headline understates the impact for most direct-to-consumer shippers because surcharges rise faster: additional handling is up 6.8% to 7.6%, residential pickup 9.2%, declared value 9.1% and extended residential delivery area surcharge 9.1%. UPS typically announces its own increase within a few weeks of FedEx.
Should you offer free shipping on everything instead of using a threshold?
Only if your gross margin on the smallest viable order still covers shipping. The floor test is shipping cost divided by gross margin: at $8.50 shipping and 55% margin, orders below about $15.45 of merchandise lose money once you absorb delivery. Always-free shipping is simplest and converts best, a threshold protects margin and lifts order value, and a flat rate is the right answer when your shipping costs vary too much by weight or zone for either to be safe.
When should you change your free shipping threshold?
Reprice it whenever your blended shipping cost moves by more than about 5%, your margin changes, or your order-value distribution shifts, and check it at least twice a year. Do not change it during peak trading: a threshold change alters both conversion rate and average order value at once, so running it in late November makes the result unreadable and the risk asymmetric. Early January, right after the carrier increases land, is the natural window.
The takeaway
The formula gets you a defensible starting number in two minutes and you should use it. Then spend an hour on the part that decides whether the policy makes money: bucket your orders, work out what you are absorbing on the carts that never needed persuading, and calculate the volume lift the threshold has to deliver. If that hurdle is 7% and your checkout already converts well, a threshold near your average order value is a subsidy with a marketing story attached. A higher line, stated early, with progress messaging tested separately, will usually beat it. Do the sums before 4 January, because the inputs change that week whether you look at them or not.