What Does a $6 Add-On Actually Leave a Plant Shop?

The box is already taped. The label is already printed. A small accessory goes into the void space beside the pot, the lid closes, and the parcel goes on the same van it was always going on. Nothing about the shipment got harder.

The six dollars that accessory just added has to survive four things before any of it reaches your bottom line: the fee base, the carrier charge, the minutes of labor it costs to pick, and the gap between the margin you quote and the margin your accounts calculate. Most shops look at one of those four.

The fee is rarely charged on the six dollars alone

Marketplace fees are calculated on more than the item price, and this is the first place the arithmetic goes soft.

On Etsy, for instance, the 6.5% transaction fee is charged on the listing price plus shipping plus gift wrapping. For sellers in the US, that 6.5% does not apply to sales tax. There is also a $0.20 listing fee per item. Etsy lists that fee as recurring on a four-month listing cycle, and how it is renewed depends on your auto-renew setting and on whether the item sold and needs relisting.

Payment processing is a separate charge with a separate base. Etsy states that the processing fee is assessed on the total amount of the sale, including tax and shipping. The rate depends on the country where your bank account is held, so it is not a number worth quoting from memory. Read your own account's current rate sheet.

Those two bases are not the same, and the difference matters in one specific way. For a US seller, sales tax sits inside the processing base and outside the transaction fee base. A shop comparing “the marketplace takes X%” against its own invoice will find the two numbers disagree, and neither is wrong. Sellers in other countries should check how their own tax treatment is handled rather than assume the US pattern applies.

The second thing to notice about the transaction fee base is what it includes. Shipping is in it. A dollar you collected and will hand straight to the carrier still gets a percentage taken out of it on the way through.

Shopify's side works differently but the principle is the same. Processing fees depend on the plan and the card type. A third-party transaction fee may also apply, but whether it does depends on your payment provider and your account settings, and some configurations are exempt from it. Shopify's own documentation describes the processing rate as varying by card type and by the location where the card was issued. Check your current plan rather than a number someone quoted two years ago.

One more thing about the fee base. On a flat-rate shipping setup, the shipping portion of the fee base does not grow when the parcel gets heavier, because the customer was charged the same amount. The item price still counts, and so does any tax that sits inside the base, so the base is not frozen. It is only the shipping component that stops rising. The extra cost drains out through the carrier bill instead. It is worth knowing which of these two shapes your own checkout has, because the fix is different.

The weight rounding rule is where accessories get expensive

Carriers do not bill by the fraction. Your item weighs what it weighs, but the rate is calculated on a rounded weight increment, and on a pound-based service that increment is a step, not a slope.

USPS is explicit in its Domestic Mail Manual: for the service in question, an item weighing 1.25 pounds is rated at a 2-pound increment. Treat that as an illustration of the rounding rule on a pound-rated service, not as a statement about every service or every weight band. Some services work in different increments, and dimensional weight can determine the charge regardless of what the parcel weighs.

Put an accessory that weighs a third of a pound into a parcel already sitting just under a tier boundary, and the whole shipment can get re-rated. Not the accessory. The shipment. The cost of that one decision can exceed the accessory's entire contribution, and it will never appear anywhere labelled as the accessory's fault.

Two recent changes are worth knowing about, with the caveat that they describe specific rate movements rather than a per-pound figure. USPS raised Ground Advantage pricing in 2026 for parcels in the 3-to-5 pound range going to Zones 5 through 8, with Zone 5 up one dollar and Zones 6 through 8 up two dollars. Any move of that size gives you a sense of the magnitude at stake. A temporary peak-season price change also ran from 4 October 2026 to 17 January 2027. Separately, a shift to destination-ZIP rate groupings took effect on 18 October 2026. That is a change in how rates are grouped by destination, not a new set of weight tiers and not a uniform increase across all USPS services. Read your own bill for the rate that applies to you.

The practical rule that comes out of this: before adding an item to the catalogue, weigh your three most common parcels as currently packed, then weigh each with the candidate accessory inside. Record the actual weight, the packaged dimensions, and the billed weight you would be charged, side by side. Do not treat the whole gap between actual and billed weight as a rounding loss. Part of that gap may be rounding, part of it may be dimensional weight, and part of it may be the rate structure of the service you chose. The point of the column is to show you which mechanism is driving your bill on the parcels you actually send, because no rate negotiation can fix a tier you are crossing unnecessarily.

Markup and margin are two different numbers wearing the same word

This is the second place money disappears without anyone noticing, and it is not a marketplace problem. It is an arithmetic one, and the cause is a single word: denominator.

Markup divides profit by cost. Margin divides profit by the selling price. When the cost is positive and the price is above that cost, the selling price is larger than the cost, so a margin expressed as a percentage is smaller than the markup that produced it. That is the normal case in a shop that prices above cost. Sell at or below cost, and the relationship does not describe what you did.

Standard management accounting states the relationship directly: selling price equals cost multiplied by one plus the markup rate. Take an item costing 140. Profit is 140 selling price is 28.57%, and the gap between 40 and 28.57 widens as the markup rises. The profit here is the difference on the cost basis you chose. It is not automatically your net profit.

A common textbook example makes the two numbers visible side by side: a retailer buys at 30, which is a 50% markup on cost and a 33.3% gross margin on price. Same ten dollars of profit. Different denominator. Both statements are accurate.

Shop owners are not being dishonest when they say “I make 40% on that.” They are quoting the number they used when they set the price. The trouble starts when that number is compared to someone else's number calculated on the other base. A supplier quoting a 20% markup and a shop reporting a 16.67% margin may be describing identical money, because a 20% markup on 120, and the $20 left over is 16.67% of the price. Round either figure and they stop being exactly the same number, which is how two people with identical books end up arguing about a percentage point.

There is a reason the two numbers get used in the two situations, and it is not carelessness. When you set a price, the cost is known and the price is still open, so you work forward from cost. When you look back at a sale, the price is known and the cost is a fact, so you work from the price. Pricing and reviewing are opposite directions through the same arithmetic, and each one naturally puts the known number in the denominator. That is why the same shop can quote 40% markup on Monday and report 28.57% margin on Friday without anyone having made a mistake on either day.

For an add-on this matters concretely. A small accessory with a high markup percentage can look like the most profitable thing in the order while contributing fewer actual dollars than a heavier item with a lower markup. Percentages do not pay rent.

Where the six dollars goes, in a hypothetical order

The following is an assumption set, not a measurement of any real shop, and every number in it should be replaced with your own. The retail price and the landed cost are chosen to make the arithmetic easy to follow. They do not represent any category's customary markup, and in everything I could find in public sources there is no citable benchmark for what garden accessories are marked up at, so do not read this as one.

Take a 2.40, sold through a marketplace charging a 6.5% transaction fee, with the shop using flat-rate shipping:

• Customer pays for the item: $6.00

• Transaction fee at 6.5%: $0.39

• Payment processing at your account's rate: depends on country and volume, so model it yourself

• Carrier charge because the parcel crossed a weight tier: $1.00

• The item's landed cost: $2.40

• Packaging and dunnage it needs of its own: $0.25

• Pick and pack time, at whatever your hour is actually worth: $0.40

On the processing line, the structure is the same almost everywhere even when the rate is not: a percentage of the total plus a small fixed amount per transaction. That structure has an unhelpful property for small items, and the degree to which it applies depends on how the add-on was bought.

If the accessory was added to a payment the customer was making anyway, the fixed per-transaction amount had already been incurred by the plant. The accessory brings the percentage charge and nothing else. If the accessory required a separate payment that you then combine into one shipment, you have created a new transaction on top of an existing one, and that extra fixed amount is a real new cost of the add-on. Which of those two happened is worth knowing, because it changes one line of the arithmetic and shops never check.

Before payment processing, what is left is $1.56. Against six dollars collected, that is 26%. Read that as a subtotal, not as the contribution. It only deducts the items listed above. Your own processing rate comes off next, and so do the costs that do not appear in this list at all: any relisting fee if this is a marketplace, the ad spend that put the item in front of the customer, and the support time an accessory quietly generates later. Fill those in before you decide anything.

Now remove the carrier line. If the same accessory fits in the void space of a box you were already sending at the same weight tier, that charge does not appear. The item leaves 1.56 before those additional costs. Nothing about the accessory changed. The parcel did.

That swing is larger than the transaction fee, larger than the packaging, and larger than the labor. It is also the line most shops never write down, because it arrives as a changed shipping cost rather than as a charge anyone can point at.

The conclusion does not depend on the two assumptions that are hardest to verify. If the item cost 2.40, or if the retail price were 6, the absolute numbers move and the ranking of the cost lines does not. The weight tier stays the biggest single variable. That is why it is worth measuring before it is worth optimizing anything else.

The add-on that fits the existing parcel is a different product decision

Two accessories with identical cost and identical retail price can leave very different amounts, depending on whether adding one changes the shipment.

An item that fits in the void space of a box you were already sending costs you its landed cost, a slice of fee, and a few seconds of picking. An item that forces a bigger box, a second box, or a higher weight tier costs you all of that plus a changed shipment. The second one has to be worth considerably more to earn the same place in the order.

This is also why the same accessory can be worth selling on one order and not on another. A small bag of labels makes sense added to a plant shipment that already has room. The same bag as a standalone order carries its own label, its own fee base, and its own minimum labor, and often should not be sold alone at all.

Contribution is not profit

Even a well-calculated add-on only earns its contribution. It covers its own cost, its share of the fee, the packaging it consumes, and the minutes it takes to pick. Those are the costs the add-on caused. The rent, the lights over the benches, and the hour spent at nine at night answering email about a plant that arrived bent are not costs the add-on caused, and contribution was never meant to cover them.

The accounting here is straightforward once the two terms are kept apart. Contribution is what is left after variable costs. Fixed costs sit below that line. Business income is total contribution across the same scope and the same period, minus the fixed costs of that period. One accessory's contribution is not comparable to the shop's net income, because the two numbers are on opposite sides of the fixed-cost line. Comparing them tells you nothing except that one number is smaller.

What the comparison can do is mislead you in either direction. A shop that adds up all its add-on contributions, finds the total smaller than monthly net income, and concludes the accessories are underperforming has compared a contribution total against a profit figure and learned nothing about either. A shop that finds the total larger has done the same thing in the other direction and may feel better for no reason.

The useful question is narrower. Did this add-on cover the extra cost it caused, and does the total contribution it belongs to cover the fixed costs of running the shop? Those are two separate tests, and only the first one is about the accessory.

What to record, if you want to know the real number

Three numbers per order will get you further than any benchmark someone quoted you, because a benchmark from another shop describes another shop's customers, prices and parcels. This article does not use an industry attachment rate, and that is a choice about what the number would be able to tell you rather than a claim that no one has ever published one.

Decide first what you are measuring, because two different figures get called attachment rate and they answer different questions. If you want to know how well the offer converts, the denominator is the orders where the customer actually saw an applicable offer. If you want to know how often orders leave with something extra, the denominator is the order pool for the period, offer or no offer. Both are legitimate. Mixing them produces a percentage that means nothing, and it is the reason two shops can quote very different attachment rates while behaving identically.

Record those orders and the add-on revenue attached to them. Then record, for the two or three accessories you sell most, the actual pick time and the actual packaging they consume. That last part is the one shops skip, and it is where the surprise usually lives.

A month of this is a starting observation, not a settled picture. A month can be a seasonal month, a month with a promotion in it, or a month where one large order dominated the total. Treat the first month as a baseline you will compare against, and give it at least one full season before you set a target from it.

A cheap add-on has to clear its own extra cost

An add-on earns its place in the parcel when it can be packed safely, when it answers something the buyer actually needed, and when what it leaves behind still covers the extra cost it caused. Not changing the box or the weight tier is the easiest way to meet the third condition. It is not the only way, and a shop that treats it as a rule will turn down accessories that would have paid for themselves.

The numbers above show why. The accessory that forced a higher tier still left $1.56 before processing and before the costs the list leaves out, and that was with a full dollar of carrier charge taken off. Change the landed cost or the retail price and the figure moves. Whether it clears your own bar depends on the bar, not on which tier the parcel landed in.

The third condition also catches the cases that look fine on the shelf. An accessory that generates a replacement shipment carries a label and a carrier charge you paid twice, and it should be charged against the accessory when you decide whether to keep it. That is a cost to deduct, not a verdict that the item is losing money, because a 4 one cannot.

The condition most often faked is the second one. A question the buyer already had is a question they were holding when they clicked the button, not one you think they ought to have. The difference shows up in the returns, not in the conversion rate, which is why it takes a season to learn and a single afternoon to ignore.

So the number to keep in your head is not the six dollars. It is the amount left after every cost this item caused: product, fees, packaging, picking, any postage it added, and the support it will generate. If that number clears what you need it to clear, the accessory stays, whatever box it arrived in. If it does not, you have just found a product that costs you money to sell, and no amount of volume will fix the direction of that arithmetic.

If you want to run this arithmetic on your own catalogue, the two inputs that are hardest to get right are the landed cost of each accessory and the packed weight of your three most common parcels. Those are the two numbers nobody publishes and every operator needs. Comparing the accessory against retail packing options before you commit is the cheapest way to find out whether it fits what you already send.

Compare the accessory cost and packing before you add a new line. Send Scarecrow Garden your shortlist and usual parcel dimensions to discuss mixed sourcing and retail packing options.