The short version: Two benchmarks, and the second is the one that matters. Your subscription should be about 1% of revenue and no more than 1.5%. Your all-in cost, subscription plus payment processing, should land between 4% and 6%. At $12,000 a year in sales that means roughly $120 in subscription and $620 all-in. If your all-in number is above 8%, something specific is wrong, and it is almost never the subscription: it is a marketplace fee, a percentage model past its crossover, or a stack of add-ons nobody has re-totalled since they were switched on.
Because the two costs behave completely differently and get confused constantly.
The subscription is fixed. It does not change with your sales, so as a percentage of revenue it falls as you grow. A $120-a-year plan is 1% of $12,000 and 0.12% of $100,000.
Processing is proportional. It scales with sales and has a floor set by card rates, so it stays roughly constant as a percentage: about 4.1% at $25 orders, 3.5% at $50.
That means the two need separate targets. A single "software should be X% of revenue" number hides the fact that one half is a decision you make once and the other is a rate you barely control.
Using a $120-a-year subscription and 2.9% plus 30¢ at $25 orders throughout.
| Annual sales | Subscription | Processing | All-in | All-in as % |
|---|---|---|---|---|
| $6,000 | $120 | $246 | $366 | 6.1% |
| $12,000 | $120 | $492 | $612 | 5.1% |
| $24,000 | $120 | $984 | $1,104 | 4.6% |
| $50,000 | $120 | $2,050 | $2,170 | 4.3% |
| $100,000 | $120 | $4,100 | $4,220 | 4.2% |
Two things fall out of that.
The all-in percentage converges on the card rate. At $100,000 it is 4.2%, which is essentially just the effective processing rate. The subscription has become a rounding error.
Below about $10,000 a year, the subscription genuinely matters. At $6,000 it is 2% of revenue on its own, which is why free and percentage-based options are competitive at the very bottom and stop being competitive quickly.
Three numbers worth writing down.
Subscription at or under 1.5% of revenue. At $12,000 in sales that is $180 a year. A $348 website builder is 2.9%, which is not automatically wrong but should be a decision rather than a default.
All-in between 4% and 6%. Anything in that band is normal for a small food business taking cards.
All-in above 8% means something is wrong. Not "you should shop around," but "there is a specific line item causing this and you can name it in under a minute once you know where to look."
Our guide to calculating the real cost per item covers where software sits against ingredients, packaging, and time, and it is usually a smaller share than vendors expect.
Five causes, in rough order of frequency, and all of them are identifiable.
A marketplace fee. Etsy charges $0.20 per listing, 6.5% of the item price including shipping, and 3% plus $0.25 in US processing: about $2.83 on a $25 order, or 11.3%. That is not a software cost, it is customer acquisition, and it is only wrong if Etsy is not actually finding you customers.
A percentage model past its crossover. A platform charging 5% plus 55¢ costs $864 a year at 40 orders a month of $25, against $144 for a flat plan. The crossover is roughly $2,000 in annual sales, so most vendors passed it long ago.
A free plan with an expensive rate. Cheddar Up's Basic tier is $0 and charges 3.95% plus 95¢, which on a $25 order is 7.8%. Free subscription, most expensive processing in the category.
Add-ons that were never re-totalled. Cococart's modules are $19 to $59 each: a store plus loyalty plus email is $77 a month, or $924 a year. Each looked small on the day it was added.
Per-location multiplication. Square Online's paid plans are priced per location, so three markets on Plus is $147 a month rather than $49, or $1,764 a year, against $0 on its own free plan.
Notice that four of those five are structural rather than a matter of shopping around. You do not fix them by finding a cheaper platform; you fix them by changing the model or the tier.
The Etsy case is the one to think about rather than react to. An 11.3% all-in on a marketplace is not automatically a problem, because you are buying discovery rather than software. Etsy's own shop statistics break traffic down by source, and the share arriving from Etsy search is what tells you whether the premium is working. If almost all your Etsy traffic is people you sent there yourself, you are paying marketplace rates to process your own customers, and that is the version worth fixing.
Four things worth checking annually.
A tier bought for a processing discount that does not pay for itself. Shopify Advanced costs $3,240 a year more than Basic and saves 0.4 points, needing about $810,000 in annual sales to break even. Square Plus at $49 a month needs roughly $12,250 a month.
A cheap plan with a hidden fee. Squarespace Basic at $19 carries a 2% store transaction fee, making it more expensive than the $29 Core plan above $500 a month in sales. The cheap plan is the expensive one.
A penalty you could avoid. Shopify charges 2% per order for using any processor other than Shopify Payments: $360 a year at $18,000 in sales, for nothing you receive.
Capability you do not use. A website builder at $348 a year when your customers arrive from Instagram already knowing what they want, or a farm platform at $1,068 when you do not sell anything by weight. Both are good products bought for a business other than yours.
Depends on your size, and the honest answer is that it should be small.
The pattern is that price matters most when you are smallest, which is the opposite of how most people shop. A vendor doing $6,000 a year should care about the $120. A vendor doing $80,000 should stop thinking about it.
Worth including, because it usually dwarfs the software.
At 30 orders a week on a manual setup, order admin runs to roughly an hour a week, or 50 hours a year. Value that at anything and it exceeds every subscription on this page combined.
So the real budgeting question is not "what should software cost" but "what does this software have to save me to be worth it?" At $120 a year, a platform needs to save you about four hours annually to break even. Almost any of them clear that in the first month.
That reframing also explains why the cheapest option is often not the right one. A $0 platform that leaves you reconciling two lists is more expensive than a $144 one that does not.
Worth naming, because software is an easy thing to optimise and rarely the binding constraint.
In rough order of return for a small food business:
Point three is the one vendors resist and it dominates everything else on this page. Our guide to why charging what you are worth feels wrong covers the psychology, and our cost per loaf breakdown shows how to work out what the number should be.
None of that argues for ignoring software costs. It argues for settling them in twenty minutes once a year and spending the rest of your attention on the four things above it. A trial that takes one real week of orders is enough to settle the software question properly, and then you can stop thinking about it.
Our guides to batch economics and cost per unit and profit margin benchmarks for food products cover the costs that genuinely move your margin, which are ingredients, packaging, and time rather than subscriptions.
Six steps, once a year, about twenty minutes.
Step two is the one people skip, and it is where the number actually lives. The advertised rate is a floor; your effective rate is higher because of the fixed per-order fee, and the gap is your average order value.
Step five catches the drift. Modules and apps are added one at a time, each looking small, and nobody re-totals until renewal. Census Bureau data on small business statistics is a reminder that most food businesses operate at a scale where a forgotten $19-a-month module is a real proportion of profit.
Software subscriptions and processing fees are generally deductible business expenses, which softens the number slightly. The IRS's Publication 535 guidance covers how ordinary and necessary costs are treated.
If your all-in is above the band and you want to see what a flat, fully published alternative costs on your own orders, Homegrown is $10 a month billed annually with 0% commission and 2.9% plus $0.30 processing stated up front, and it handles pickup at each place you sell with its own schedule and cutoff, local delivery with a radius and a route, and sales tax calculated, filed, and remitted in all 50 states. At $12,000 in annual sales that works out to about $612 all-in, or 5.1%. The honest bounds: it does not ship nationally, has no point-of-sale, no app ecosystem, and no free tier, so under about $6,000 a year a free plan is genuinely cheaper and you should use one. You can run one real week through a trial and calculate your own percentage rather than using anyone's example.
Two numbers: the subscription should be about 1% of revenue and no more than 1.5%, and your all-in cost including payment processing should land between 4% and 6%.
At $12,000 in annual sales, roughly $120 in subscription and about $612 all-in. At $24,000, the same $120 subscription and about $1,104 all-in, which is 4.6%.
Because the subscription is fixed while processing is proportional. At $100,000 in sales the subscription is a rounding error and your all-in is essentially just the card rate, around 4.2%.
Something specific is causing it, and it is nearly always one of five things: a marketplace fee, a percentage model past its crossover, a free plan with an expensive rate, forgotten add-ons, or per-location pricing.
Only when you are very small. Under $10,000 a year the subscription is 1% to 2% of revenue and worth optimising. Above $30,000 it is under 0.5% and you should buy on capability instead.
Your time. At 30 orders a week, manual order admin is roughly 50 hours a year, which exceeds every subscription discussed here. A $120 platform needs to save about four hours annually to break even.
Once a year, taking about twenty minutes. List every recurring charge, add last year's actual processing fees, divide by sales, and check every add-on against whether you used it in the last three months.
Two benchmarks. Subscription at about 1% of revenue, capped around 1.5%. All-in, including processing, between 4% and 6%. At $12,000 in annual sales that is roughly $120 and $612.
If your all-in is above 8%, do not go shopping for a cheaper platform. Name the line causing it, because it will be one of five things and four of them are structural: a marketplace fee, a percentage model past its crossover, a free plan with an expensive card rate, add-ons nobody re-totalled, or per-location pricing multiplying a subscription you thought was $49.
And remember which direction this cuts. Price matters most when you are smallest. At $6,000 a year the subscription is 2% of revenue and worth arguing about. At $80,000 it is under 0.2%, and the hour a week you spend on order admin costs you more than the software ever will.
