The seller had two screens open. One showed TikTok Shop affiliate commission data for the month. The other showed what actually reached the bank. The gap was a little over three thousand dollars, and the first instinct was that the platform had made an error.
It had not. The two screens measure different things. Neither is wrong. Reading them as if they should agree is what turns a reporting nuisance into a rate decision made on the wrong number.
This is written for sellers running ten to fifty SKUs with one to three people on operations, where the same person sets commission rates, reads the sales report, and answers to the owner. If a finance team reconciles your numbers, half of this is already solved. The other half is what you do before changing a rate.
Both numbers are correct, which is the problem
The sales view answers one question: what did customers pay. The payout view answers another: what survived platform fees, shipping, refunds and returns. Those two questions have different answers by design.
Most teams treat the first number as the truth and the second as an error. That assumption holds until the gap grows large enough to notice, which is usually the month someone proposes cutting a commission rate.
A gap is not a failure signal. It is a signal that a gross measure is being compared against a net one. Three numbers explain almost every gap, and they need to sit together for the same product and the same period.
| Number | What it counts | Why it differs from the payout |
|---|---|---|
| Gross sales | What customers paid, before any deduction | Includes orders that will later be refunded or returned |
| Net sales after refunds | Gross minus refunds, returns, platform fees and shipping | Closest to reality, but still moves as late refunds land |
| Payout | What reached the bank, after commission | Lags by the settlement window, and converts currency on its own date |
Why a hand-built sheet cannot hold this
Every seller who sees the gap for the first time builds a spreadsheet. It works for one month. Then it quietly stops being true, and nobody notices because the totals still look plausible.
Three things break it. Refunds arrive after the period closes, so a sale booked in October and returned in November has to move between two files that were already signed off. Fees differ by product, so a single blended percentage hides the SKU that is losing money. And a second store prices in a second currency, which converts on the settlement date rather than yours.
A spreadsheet answers a question once. Reconciliation is a question you ask every week, on data that keeps arriving. That difference, not the arithmetic, is why the sheet gets abandoned around week five.
The other cost is attention. Someone has to re-export, re-paste and re-check three sources before anyone can say whether last month was good. In a team of one to three people, that task loses to whatever is shipping that week. Multiply it by a second store and a second currency, and the answer arrives late enough that the decision has already been made on a feeling.
Where the three numbers actually live
The reason the comparison is hard is that gross, net and payout normally live in three places. Making them one place is the whole job.
DAMI reports product sales performance with refunds, platform fees and shipping already deducted, so net contribution per product exists as a starting value rather than something you derive. The same view drills into SKU sales detail, which is how you find the product a creator moved well against the one that came back. And because every store reports into one workspace, a rate set in one shop is compared against the same deductions in the next.
That changes the weekly routine from rebuilding a sheet to reading one. Filter to the store, pick the period, sort by net contribution. The question of whether a creator is worth their rate then has an answer that took a minute instead of an afternoon.
The affiliate commission data on the creator side matters as much as the product side. A creator can look strong on a blended margin and weak on the one product that funds the campaign, and only a per-creator view of the same deductions shows which one you are looking at.
DAMI puts both halves on one record: what the creator moved, and what the same orders cost after refunds and fees. Read together, the rate conversation stops being about the number the creator asked for and starts being about the margin that number has to survive.
For the mechanics of which deductions belong above the line, net sales after refunds walks through them. It is worth reading before trusting any total, including ours.
What to look at before you cut a rate
Rate changes get decided quickly, usually after a month where the payout looked weak. The data that would justify the change almost always exists, and almost never gets assembled first. By the time the figures are ready, the meeting has usually happened and the new rate is live.
Three figures need to be on one screen for the specific creator. Net contribution per product after refunds and fees. Posted video count over sixty days. And the share of that creator’s sales that came back as returns.
One honest limit. No tool can tell you what a refund will do next month, and the settlement window means the most recent weeks are always incomplete. DAMI shows what has landed, not what is coming. Treat the newest month as provisional and never cut a rate on it alone.
The second limit is attribution. When a creator’s video runs while a paid campaign is live, the sales record cannot separate the two. A decision assuming all of it came from the creator will be wrong one way, and one assuming none did will be wrong the other. Neither number exists. Say so before the meeting, because someone will ask.
For what a healthy rate looks like once the numbers are clean, commission optimization covers the range each category can carry. The DAMI sales and commission workspace is where net contribution and the commission rate sit on the same line.
FAQ
Why does my affiliate commission data change after I export it?
Because refunds, returns and currency conversion keep landing after the period closes. An export taken on the first of the month is a snapshot of an unfinished month. Take it again a week later and the difference is late activity, not an error.
Should commission be calculated on gross or net sales?
Net. Commission paid on gross means paying for orders that come back, while the creator keeps the difference. Negotiation also gets easier once both sides look at the net figure, because the gap stops being a surprise halfway through a campaign.
How large a gap between report and payout is normal?
There is no universal figure, and anyone quoting one is guessing about your category. What matters is that the gap is stable and explainable. A stable gap is arithmetic. A gap that widens two months running is a product problem, usually returns, and worth investigating before touching commission at all.
