The short version: Cost per order is the only unit that lets you compare a subscription against a percentage against a marketplace fee, because it collapses all three into one number. The formula is (annual subscription ÷ annual orders) + (rate × average order) + fixed fee. At 480 orders a year of $25, a $120 flat plan costs $1.28 an order and Square Online's free plan costs $1.13, because the subscription spread across too few orders outweighs Square's higher rate. Those two cross over at exactly 1,200 orders a year, or 100 a month, and above that the flat plan wins permanently.
Three terms, and the first is the one everybody leaves out.
Cost per order = (annual subscription ÷ annual orders) + (percentage rate × average order value) + fixed per-order fee
Worked on a $120-a-year plan at 2.9% plus 30¢, with 480 orders a year averaging $25:
That single number can now be compared against anything: a free plan with a higher rate, a percentage model, a marketplace, or a platform that charges per location.
One number, three inputs, and every platform on your shortlist becomes comparable on the same line. Our break-even guide for flat against percentage pricing picks up where this calculation ends.
Because it is the only one that makes structurally different models comparable.
Compare a $144-a-year subscription against a 5%-plus-55¢ model and you are comparing a fixed number to a rate. Convert both to cost per order and they are the same kind of thing.
It also matches how you actually think about a business. You know roughly how many orders you take and what they are worth. You do not naturally think in annual fee-bearing revenue.
And it exposes the thing percentage pricing hides: your subscription cost per order falls every time you take another order, which is why the same platform can be expensive at 10 orders a month and cheap at 100.
At 480 orders a year averaging $25, which is $12,000 in sales. All figures from each company's own published pages in July 2026.
| Platform | Subscription | Free trial | Platform fee | Card processing | Cost per order |
|---|---|---|---|---|---|
| Hotplate, fee passed on | $0, no subscription | no free trial needed, nothing to subscribe to | 5% + $0.55, charged to your customer | 2.9% + $0.30, paid by you | $1.03 |
| Square Online Free | $0 | no free trial needed, free forever | $0, no commission | 3.3% + 30¢ online | $1.13 |
| Homegrown | $10/mo billed annually | 7-day free trial | $0, 0% commission | 2.9% + $0.30 | $1.28 |
| Big Cartel Platinum | $12/mo billed annually | 7-day free trial | $0, no commission | rate not set by Big Cartel: your own processor, commonly 2.9% + 30¢ | $1.33 |
| Bake.Shop | $149/yr annual | 14-day free trial | $0, commission-free | 2.9% + 30¢ via Stripe | $1.34 |
| Cheddar Up Basic | $0 free forever | no free trial needed, free forever | $0 platform fee | 3.95% + $0.95 | $1.94 |
| Hotplate, fee absorbed | $0, no subscription | no free trial needed, nothing to subscribe to | 5% + $0.55, absorbed by you | 2.9% + $0.30, paid by you | $2.83 |
| Etsy | $0; Etsy Plus optional $10/mo | no free trial: listings are pay-per-item | $0.20/listing + 6.5% of item price | card processing 3% + $0.25 via Etsy Payments | $2.83 |
The spread between the cheapest credible option and the most expensive is $1.80 an order, which across 480 orders is $864 a year. That is a real difference, and it is almost entirely explained by fee structure rather than by which company you chose.
The ranking flips, and knowing where is the useful part.
Comparing Homegrown at $120 a year with 2.9% plus 30¢ against Square Online Free at 3.3% plus 30¢, on $25 orders:
The crossover is 1,200 orders a year, or 100 a month. Below that, a free plan with a higher rate is genuinely cheaper. Above it, paying $120 to get 0.4 points off is the better deal, and the gap widens slowly.
You can calculate your own crossover directly: annual subscription ÷ the rate difference × average order value. Here that is $120 ÷ (0.004 × $25) = 1,200 orders.
Because the fixed fee is a bigger share of a small order, and the percentage is a bigger share of a large one.
At the same 480 orders a year, on a $120 plan at 2.9% plus 30¢:
So the same platform costs you 8.4% or 3.5% depending purely on basket size. The conclusion is the same one that falls out of every fee calculation in this category: raising average order value is the biggest lever you control, because the fixed 30¢ is 1.2% of a $25 order and 0.3% of a $100 one.
It also changes which platform wins. A high fixed fee like Cheddar Up's 95¢ is punishing at $10 and unremarkable at $100. A high percentage like Etsy's 6.5% is the reverse.
So there is a genuine strategy here rather than just a comparison. If your baskets are small, hunt for a low fixed fee. If your baskets are large, hunt for a low percentage, and look at whether bank transfer is available, since that is where the large-order saving actually lives. Our guides to minimum order values for food delivery and profit margin benchmarks cover moving basket size deliberately, which changes the fee answer as well as the revenue one.
Six steps, about ten minutes with last year's numbers.
Step five is the one that keeps this honest. A 15¢ difference sounds trivial and is $72 a year at 480 orders. A 90¢ difference is $432. Neither is life-changing and both are worth ten minutes.
Step six prevents switching twice. If you are growing and the ranking flips at 1,200 orders, and you are currently at 900, you will be on the wrong side of the crossover within a year.
There is a fair objection to all of this: a difference of 15¢ an order is not worth a weekend of migration and a fortnight of customer confusion. That is correct, and it is the right conclusion when the gap is small. The calculation earns its keep in the cases where the gap is not small: a marketplace at $2.83 against a storefront at $1.28, or an absorbed percentage fee doing the same. Those are $744 and $864 a year respectively, and they are worth acting on.
So use the number to sort your options into three groups: close enough to ignore, worth switching for, and worth fixing without switching, which is where a tier change or passing a fee on usually lands. Running one real week through a trial is what turns the third group from a guess into a measurement.
Include everything that varies with selling, and be honest about it.
Include:
Exclude:
That last exclusion is deliberate rather than dismissive. Time is usually the largest cost in a small food business, and mixing it into a per-order fee comparison makes both numbers useless. Calculate the fee comparison cleanly, then ask separately whether the more expensive platform saves you hours.
Done in that order, the answer is often that the pricier option is obviously correct. A platform costing 15¢ an order more is $72 a year at 480 orders. If it removes an hour a week of reconciliation, that is 50 hours for $72, which nobody would turn down if it were presented that way. The mistake is comparing fees and hours in the same number, where the fee difference looks large and the hours disappear entirely.
Our guides to calculating the real cost per item and batch economics cover the product-cost side, where the numbers are considerably larger.
Three things, and they are the reason to do the calculation at all.
Whether a free plan is actually free. Square Online Free is genuinely cheaper below 1,200 orders a year and genuinely more expensive above it, so the answer changes with your size rather than being fixed. Cheddar Up Basic is $0 and costs $1.94 an order, which is more than several paid plans on this list.
Whether a marketplace fee is buying anything. Etsy at $2.83 an order is 2.5× a storefront. That premium is customer acquisition, and it is only worth it if Etsy is finding you customers you would not otherwise have.
Where your own crossover sits. Not a general rule but your number, based on your own order count and average basket. That figure is the thing to write down and check once a season, because it moves as you grow and nobody will tell you when you have passed it.
The IRS's recordkeeping guidance covers what you should be retaining anyway, and a year of order counts and fee totals is exactly what makes this calculation take ten minutes rather than an afternoon. The SBA's guidance on managing your finances is a reasonable framework for keeping those records in a form you can actually use.
If you want to run the numbers on a platform with the rate published rather than discovered, Homegrown is $10 a month billed annually with 0% commission and 2.9% plus $0.30 processing stated up front, which works out to $1.28 an order at 480 orders of $25 and $1.08 at 2,400. The honest bounds: it does not ship nationally, has no point-of-sale, no drop windows, and no free tier, so below roughly 100 orders a month a free plan is cheaper on this arithmetic and you should use one. You can run one real week through a trial and calculate your own cost per order from actual orders rather than assumed ones.
Divide your annual subscription by your annual order count, then add the percentage rate times your average order value, then add the fixed per-order fee. That total is directly comparable across any pricing model.
Because it makes structurally different models comparable. A subscription, a percentage, and a marketplace fee are three different kinds of number until you convert them all into what one order costs.
Against Square Online's free plan, a $120-a-year plan at 2.9% wins above 1,200 orders a year, or 100 a month. Calculate your own with: subscription ÷ (rate difference × average order value).
Yes, substantially. A high fixed fee like Cheddar Up's 95¢ is punishing on $10 orders and unremarkable on $100 ones, while a high percentage like Etsy's 6.5% works the other way round.
No, keep it separate. Time is usually the largest cost in a small food business, and mixing it into a fee comparison makes both numbers useless. Compare fees cleanly, then ask whether the pricier platform saves hours.
At $25 orders, roughly $1.03 to $1.35 on the credible options, which is 4% to 5.4%. Above $1.90 an order you are on a marketplace, a percentage model, or a free plan with an expensive card rate.
Once a season, and always before switching. Check whether the winner changes at twice your current volume, since that is how vendors end up switching twice inside a year.
One formula: (annual subscription ÷ annual orders) + (rate × average order) + fixed fee. It is the only unit that lets you compare a subscription against a percentage against a marketplace fee, because it turns all three into the same kind of number.
At 480 orders a year of $25, the credible options run $1.03 to $1.35 an order, and the expensive ones run $1.94 to $2.83. Across a year that spread is $864, and it is explained by fee structure rather than by brand.
The number worth writing down is your own crossover: subscription ÷ (rate difference × average order value). Against a free plan charging 0.4 points more, a $120 subscription pays for itself at 1,200 orders a year. Below that, take the free plan. Above it, take the subscription, and check once a season rather than once you happen to wonder.
