How Results are calculated

The Results page answers one question: what did the actions you deployed actually earn you?

Not "how much money moved through them" — that is a different question with a different answer, and the page keeps the two apart on purpose. This is where each figure comes from.

Shopee only

The attribution service behind this page is Shopee-only. A TikTok or Lazada shop has no ledger yet, so the page says so rather than drawing an empty chart that would read as "your actions earned nothing".

Profit credited

The headline figure. Every deployed action is measured against what would have happened without it, not against zero.

• Budget changes — modelled profit at the spend that was actually delivered, minus modelled profit at the baseline spend policy. The difference is what the change is credited with.
• Target ROAS changes — measured against the campaign's own prior rate.
• Other families — each carries its own stated basis, shown on the (i) beside its row in Where the profit came from.

Only settled money counts. Shopee clears an order roughly one to two weeks after it is placed, so the most recent days in any window read low by design and fill in later.

An action that has been deployed but has not settled yet shows as measuring, never as zero. Those are two different facts and the page never collapses them.

Ads optimizer

The first card at the top. It is the profit DataGlass earned by moving budget between your campaigns day by day, measured the same way as the budget changes above, over the same dates. It is part of Profit credited, not added on top, so the card also says what share of the total it makes up, and how much of it came through the ad groups listed further down. A figure DataGlass cannot stand behind yet reads Not measured, never zero.

Time saved

An estimate, and labelled as one. Every time the ads optimizer reviewed and changed your budgets counts as one budget review: once for each day it set them, and once more for each later time that day it changed them again. Only changes Shopee accepted count, and changes you made yourself do not. Each review is valued at 12 minutes, the median time our own team took per budget review on the shops we manage — with direct access to the data, so doing it by hand in Seller Center usually takes longer. The card also shows how many budget changes that came to, and across how many campaigns.

Profit by day

Genuinely daily. The backend stores one row per action per day and refuses a series that does not sum back to the headline figure — nothing here spreads a window total across days.

• Bars — what each day earned. A day measured at a loss is held back rather than shown, so it has no bar and reads Not shown yet; the running total stays flat across it.
• Line — the running total across the whole period.

Switch to Table for the same days as figures: what each day earned, the running total after it, and how many actions were measured that day, newest first.

If a family is measured over a window rather than per day, its money is in the headline but not in the bars. The page names that gap under the chart instead of leaving a chart that silently fails to add up.

Where the profit came from

The same total, split by action family. Each family has its own measurement basis, which is why the split matters more than the ranking — the (i) on each row states the basis for that row.

Percentages are shares of the earning families only. A family measured at a loss is held back and reads Not shown yet; it is not a slice of anything. The number of actions you deployed in the window sits beside the heading.

Ad groups DataGlass runs

This table carries two different bases side by side, and they must never be added.

Measured uplift is the last column because it is the only one that is ours. It is what the budget writes added; everything to its left is what flowed through the campaigns whether we ran them or not.

Shopee's ad attribution answers "what ran through these campaigns". Measured against items, that same basis has claimed between 104% and 218% of an item's entire sales as ad-driven — which is why it sits beside the measurement as context and is never summed into it.

Two further limits, both from the platform rather than from choice:

• The group endpoint takes no date range, so the attributed columns cover the shop's ads reporting window while the uplift column obeys your date picker. The two columns legitimately span different periods — and the attributed one predates DataGlass, covering the campaigns from before the group was ever handed to us. It is not a post-management figure and must not be read as one.
• There is no group-level daily endpoint. A shop-wide daily chart would mean one request per campaign, so day-by-day figures live on each group's own page instead.

A group reads Not measured until its budget writes are mature and its campaigns priced. That is not zero.

Worth reviewing

• Worth a look — measured, and currently negative over the action's own window. A loss this early can still turn around.
• Measuring — deployed, but without enough settled orders to report yet. A long list here is normal on a recent deploy.

Why a figure may not match Seller Center

Seller Center counts every order placed, including cancelled, unpaid and in-transit. This page counts settled money attributed to a specific action. They are different questions, and a gap between them is usually a definition difference rather than missing data.