Running a Real Invoice Through the Fee Waterfall
1,647 words · about 8 minutes · note 36 of 40 · Step 7, Step 8
An invoice total tells you what left your balance; it does not tell you which line grew. The fee waterfall on this site’s ledger page exists to answer the second question, by splitting one order into nine named lines from merchandise to destination handling and showing what share of the total each one holds. The sections below walk the tool the way it is actually used: what appears when it opens, which fields start with a value and which start empty, how to read the output without over-reading it, and a single invoice carried through the whole run so the arithmetic can be checked line by line.
What appears when the waterfall opens
The waterfall sits at the top of the ledger page, above the rate and transit tables. It is a single column of nine lines in the order money leaves your account: merchandise, domestic freight, service fee, payment fee, quality-check photos, merge and reinforcement, international postage, a tax estimate, and destination handling. Each line carries an amount, a one-line basis, and a marker for whether the line is controllable — that is, whether a decision of yours moves it.
Nothing on the page is a quotation. The tool multiplies numbers you supply, and every register value behind it is marked not verified because no live rate could be confirmed at review. Reading it as an estimate of your own inputs is the correct frame; reading it as a price is not.
The controllable marker is the part of the interface worth understanding early. Five lines carry it — service fee, payment fee, quality-check photos, merge and reinforcement, and international postage — because each responds to a decision: which options you accept, how many photo requests you make, whether parcels are combined, and which line is booked. Merchandise, domestic freight, tax and handling do not, because they follow from a seller, a rule or a carrier rather than from a choice made at the packing bench. Sorting the nine lines that way turns the waterfall from a receipt into a list of levers.
Nine lines, one weight setting, and the defaults behind them
Two fields open empty on purpose: merchandise and domestic freight. Merchandise is the figure you carry in from your own rows, and domestic freight is quoted per seller and per basket, so no register value could stand in for it without inventing one. Everything else opens with a template value that you can overwrite.
The percentage lines attach to merchandise only. That is the detail most often misread: a six percent service fee and a three percent payment fee are calculated on the goods value and never on postage, so a heavier parcel does not enlarge them. Fixed lines — the photo request, the merge and reinforcement charge, handling — attach per request or per parcel instead.
| Waterfall line | Register default | Unit | What it attaches to | Status |
|---|---|---|---|---|
| Merchandise | none | USD | seller price | entered by you |
| Domestic freight | none | USD | seller to warehouse | no default held |
| Service fee | 6 | % of merchandise | goods value only | not verified |
| Payment fee | 3 | % of merchandise | goods value only | not verified |
| QC and photos | 2 | USD per request | each request | not verified |
| Merge and reinforcement | 1.5 + 1 | USD per parcel | each parcel | not verified |
| International postage | 11 | USD per kg | chargeable weight | not verified |
| Tax estimate | 0 | % of merchandise + postage | goods plus postage | not verified |
| Destination handling | 0 | USD per parcel | each parcel | not verified |
| Volumetric divisor (setting) | 6000 | g/L | volume ÷ divisor × 1000 | not verified |
Where a default is a placeholder rather than an estimate
Three of the defaults behave differently from the rest, and treating them alike is how a plausible-looking total goes wrong. The divisor is a setting rather than a charge: it changes how many kilograms the postage line is multiplied by, so it should be set from the line you intend to use, not left at whatever the field opened with.
Tax and destination handling both open at zero. That zero is a deliberate refusal to guess: duty and VAT depend on destination, category and declared value, and a handling fee depends on the carrier presenting the parcel. A total that includes a zero tax line is a floor, not a forecast, and it should be read that way until you have a rate to type in.
The per-kilogram postage field expects one effective number. If your line publishes a first-kilogram band plus a continuation band, convert it before typing: a band of 18 for the first kilogram and 11 for each further kilogram becomes an effective 20 per kilogram at 2 kg, because 18 + 22 = 40 and 40 ÷ 2 = 20.
Reading the output: shares before totals
The total is the least informative number the tool produces. Shares are where the decisions are: a line holding four percent of the total is not worth an hour of optimisation, and a line holding a fifth of it is worth a different choice of route or a different packing decision.
Two readings follow from that. Look for the largest controllable line first, since merchandise and postage are usually the two biggest entries and only one of them responds to a decision you make after the order. Then look at the ratio between the non-goods lines and merchandise: that figure, not the grand total, is what changes when you switch line, merge parcels or drop a photo request.
One caution about the shares. They are shares of the total as currently entered, so a field left at zero inflates every other share. Filling in an unknown fee later moves the percentages even though nothing about the order changed.
Comparing two runs is where the page earns its place. Change one field, keep the other eight, and the movement in the total belongs to that field alone — switch the postage rate and the difference is pure line effect, add a second parcel to the merge line and the difference is a packing decision. Changing several fields at once produces a number that cannot be attributed to anything, which is the situation the waterfall was built to avoid.
One invoice through the waterfall
A worked run makes the arithmetic visible. Take a basket of $150.00 of merchandise, which at the register’s default rate of 6.232 CNY per USD corresponds to ¥934.80. For scale, four listings at the catalogue median of $36.18 come to $144.72, so this is a realistic small haul rather than a hypothetical one. Chargeable weight is set at 2 kg, the postage field at the template rate of $11.00 per kilogram, and the two unknown lines are left at zero.
The bar that matters is at the bottom of the table: the non-goods lines come to $40.00 against $150.00 of merchandise, an uplift of 26.7 percent, and the three largest lines account for $181.00 of the $190.00 total while the remaining six lines account for $9.00 between them.
| Line | Amount (USD) | Share of total | Basis in this run |
|---|---|---|---|
| Merchandise | 150.00 | 78.9% | basket value entered |
| Domestic freight | 0.00 | 0.0% | no default held, left empty |
| Service fee | 9.00 | 4.7% | 6 percent of merchandise |
| Payment fee | 4.50 | 2.4% | 3 percent of merchandise |
| QC and photos | 2.00 | 1.1% | one request |
| Merge and reinforcement | 2.50 | 1.3% | 1.50 + 1.00 per parcel |
| International postage | 22.00 | 11.6% | 2 kg × $11.00 per kg |
| Tax estimate | 0.00 | 0.0% | no rate supplied |
| Destination handling | 0.00 | 0.0% | unknown fee not invented |
| Total | 190.00 | 100.0% | sum of the nine lines |
Questions that come up while filling the lines
Whether postage belongs in the base for percentage fees is the most common question, and the answer in this tool is no: the service and payment lines are calculated on merchandise alone. That keeps them stable when a parcel gets heavier, which is exactly why an increase in the total after a packing change comes from the postage line rather than from the fees.
A second question concerns first-kilogram pricing. Converting a band into an effective rate, as shown above, raises this run from $190.00 to $208.00 — the difference is exactly the $18.00 fixed band, and it is worth typing both figures once to see the size of it. A third concerns the divisor: changing it from 6000 to 5000 does nothing at all until the parcel is bulky enough for volume to exceed actual weight, at which point it silently rewrites the postage line.
The fourth is about settlement. Amounts that a seller has already collected, such as VAT charged at the point of sale on a low-value consignment, do not belong in the tax line, because they never appear on the invoice this tool is modelling. Counting them twice is the easiest way to make a total look alarming for no reason.
A fifth question concerns negative or empty fields. Blank is not zero: an empty merchandise field stops the calculation, because everything downstream is a percentage of it, while an empty handling field simply contributes nothing. Where a fee is genuinely unknown, leaving the field empty and reading the total as a floor is more accurate than typing a guess that will never be corrected.
What the waterfall cannot decide for you
The tool cannot know what you paid per item, which line you will use, or what a border will charge. What it can do is keep those three uncertainties in separate lines so that an assumption is visible instead of buried in a total. A run in which tax and handling both sit at zero is an honest run; a run in which they carry invented numbers is worse than no run at all.
For scale, this site’s 218-listing snapshot taken on 2026-09-29 has a median listing price of $36.18, a p25 of $22.53 and a p75 of $53.32, so a two-row basket priced at the median lands near $72 while a four-row basket priced at the p75 lands near $213 — that gap is the range the merchandise field usually occupies for a small haul, before any fee is added. The currency tool behind the conversion is editable and loads the snapshot’s implied 6.232 CNY per USD; replace that with the rate you actually see, because the value belongs to the day you order rather than to the day of this review, 2026-09-29.