The short version: At $2,000 a month across 80 orders of $25, a 5%-plus-55¢ platform fee comes to $1,728 a year. Add the card processing you pay separately and absorbing that fee costs you $2,712 a year, against $1,104 on a flat $120 plan. The gap is $1,608. Per order, that is $2.83 versus $1.15. The decision that moves this number is not which platform you choose, it is who pays the fee: passed to your customers, your own cost drops to $984 and the platform becomes the cheapest option on the page.
Worked in full, because the components matter.
At $2,000 a month, 80 orders of $25, so $24,000 a year across 960 orders:
The 5% plus 55¢ platform fee:
Card processing you pay regardless, at 2.9% plus 30¢:
So the three scenarios:
| Arrangement | Your annual cost | Per order |
|---|---|---|
| Flat $120 plan + 2.9% and 30¢ | $1,104 | $1.15 |
| 5% + 55¢ passed to customers + your processing | $984 | $1.03 |
| 5% + 55¢ absorbed by you + your processing | $2,712 | $2.83 |
Absorbing the fee costs $1,608 a year more than a flat plan. Passing it on makes the same platform $120 a year cheaper than the flat plan.
Same platform. Same volume. A $1,728 swing on one setting.
Because it is a fixed fee on small orders, and food orders are small.
On a $25 order, 55¢ is 2.2% on its own. Add the 5% and the effective platform fee is 7.2%, not 5%.
At different basket sizes:
So the advertised 5% is the truth only on large baskets, and food baskets are not large. A vendor with a $12 average order is paying an effective 9.6%, nearly double the headline.
That also means raising average order value moves this number more than anything else you control, which is the same conclusion every fee calculation in this category reaches.
Worth converting into something concrete, because a percentage does not feel like money and this does.
At $1,608 a year you could cover:
That last one is the useful comparison, and it is worth sitting with. The fee difference over a single year buys more than a decade of the alternative subscription, which is not a rhetorical flourish but simple division.
None of that means the platform is bad value. It means the fee is a real cost with real alternative uses, and it deserves a deliberate decision rather than a default setting.
The decision worth $1,728 a year at this volume, and it is genuinely situational.
Passing it on works when:
Absorbing it works when:
Most platforms do not offer the choice at all. Hotplate adds its 5% plus 55¢ to the customer's total by default. Cheddar Up lets you configure who pays. Almost everything else simply deducts from you.
Our guide to why charging what you are worth feels wrong covers the psychology, which is usually the real obstacle rather than the arithmetic.
Test it, because the answer is measurable and the stakes are $1,728.
Step four is the one people get wrong. A pass-through version that converts 8% worse can still leave you better off, because you kept $1.80 on every order that did complete. Work out the net, not the rate.
At 80 orders a month, an 8% drop is roughly 6 orders. Those 6 orders at $25 are $150 of revenue, against $144 saved in fees on the 74 that completed. Close to a wash, which is exactly why this needs measuring rather than assuming.
Different structure, different justification, worth separating.
Etsy at $24,000 a year across 960 orders charges $0.20 per listing, 6.5% of item price, and 3% plus $0.25 in US processing. That is about $2.83 per order, or $2,717 a year, which is almost exactly what absorbing a 5%-plus-55¢ fee costs.
The difference is what the money buys. Etsy's 11.3% includes search traffic: people who did not know you existed. A drop platform's 7.2% does not, and does not claim to; it buys countdowns, reserved inventory, waitlists, and SMS.
So the honest test differs by category:
Both fees are defensible. Neither is defensible if you are not using the thing you are paying for.
Yes, and most vendors never adjust for it, which is the quiet version of absorbing it twice.
If you absorb a 7.2% effective platform fee plus 4.1% in card processing, roughly 11.3% of every order leaves before you have paid for ingredients. A price set before you were on that platform is now delivering meaningfully less margin than it did.
Two reasonable responses:
The version that fails is doing neither and treating the fee as a cost of doing business. It is a cost of doing business, and costs of doing business belong in your prices. Our guides to communicating a price increase to regular customers and minimum order values for food delivery cover doing both without losing anyone.
Worth doing the arithmetic on your own numbers rather than mine, since the answer depends entirely on your average basket. Running a month through a flat-fee trial alongside your current setup gives you both totals and makes the pricing decision concrete rather than theoretical.
Below the crossover, which is lower than most vendors expect.
Against a $144-a-year flat plan, a 5%-plus-55¢ model at an effective 7.2% breaks even at:
$144 ÷ 0.072 = $2,000 in annual sales
That is $167 a month. Above that, the flat plan wins, and at $2,000 a month you are twelve times past it.
There is one genuine exception: months when you sell nothing cost you nothing. A vendor trading four months a year on a percentage model pays only in those four months, while a subscription runs all twelve. That option value is real and it is the honest argument for percentage pricing.
At $2,000 a month, though, that argument does not apply. You are trading, and consistently.
The gap widens, which is the property of percentage pricing that catches people out.
Doubling to $4,000 a month, so $48,000 a year across 1,920 orders:
The flat plan's cost barely moved because only the processing scales. The percentage model's cost doubled exactly in step with your sales.
That is the structural point. A percentage fee is a permanent tax on growth, and every improvement you make to the business, more orders, higher prices, a second market, hands a proportional share to the platform. A flat plan takes the same $120 whether you sell $12,000 or $120,000.
Which does not make percentage models wrong. It makes them a starting structure rather than a permanent one, and the sensible response is a pre-registered switch trigger rather than an annual re-decision. Write down the monthly sales figure at which you move, and check it once a season. If you want a concrete comparison to check it against, running one month of real orders through a flat-fee trial gives you both totals rather than one and an estimate.
Four steps, in order of how much they are worth.
Step one first, because it costs nothing to establish and is larger than anything else on this page.
Our guides to minimum order values for food delivery and profit margin benchmarks cover step four, which quietly improves this number and your margin at the same time.
The SBA's guidance on managing your finances is a reasonable framework for tracking what this actually costs you month to month, and platform fees are generally deductible business expenses, which the IRS's Publication 535 guidance covers.
If you have decided the mechanics are not what you need and a flat plan is the better fit at this volume, Homegrown is $10 a month billed annually with 0% commission and 2.9% plus $0.30 processing published up front, which at $2,000 a month works out to $1,104 a year all-in, or $1.15 an order. 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. The honest bounds: there are no drop windows, no countdowns, no reserved inventory, and no waitlists, plus no national shipping and no point-of-sale. If your releases genuinely sell out and those mechanics are converting, a drop platform is doing something this does not and 7.2% may be the right price for it. You can run one release through a trial and compare on your own numbers.
$1,728 a year once the 55¢ per-order component is counted, across 960 orders of $25. Add the card processing you pay separately and absorbing it totals $2,712 a year.
A $120-a-year flat plan with 2.9% plus 30¢ costs $1,104 all-in at the same volume. Absorbing the percentage fee costs $1,608 more; passing it to customers costs $120 less.
Because of the fixed 55¢. On a $25 order that is 2.2% on its own, making the effective platform fee 7.2%. On a $10 order it is 5.5%, giving 10.5%.
Test it rather than assume. Run one release each way with the same products and compare net revenue, not conversion rate. An 8% drop in orders can still leave you better off.
Around $2,000 in annual sales against a $144 plan, using the 7.2% effective rate. At $2,000 a month you are roughly twelve times past that crossover.
No. Etsy's roughly 11.3% costs about the same per order as an absorbed 5%-plus-55¢ fee, but it includes search traffic. A drop platform's fee buys mechanics, not customers.
Establish whether you can pass the fee on, which is worth $1,728 a year here. After that, raise average order value: the 55¢ is 2.2% at $25 and 1.1% at $50.
At $2,000 a month, a 5%-plus-55¢ platform fee is $1,728 a year, and the effective rate on $25 orders is 7.2% rather than 5% because the fixed component does not scale.
Absorbing it puts your all-in cost at $2,712 a year, or $2.83 an order, against $1,104, or $1.15, on a flat plan. That $1,608 gap is a season of market fees, 65 hours of paid help, or twelve years of the alternative subscription.
The lever is not the platform, it is the setting. Passing the fee to customers drops your own cost to $984, which makes the same platform the cheapest option here. That decision is worth $1,728 a year, it is testable in two releases, and most vendors have never tested it.
