A campaign dashboard shows one impression total. The downloaded placement report adds up to something smaller. It is tempting to repair the spreadsheet until the figures agree. That can turn a visible reporting limitation into an invisible analytical error.
For teams buying TikTok Ad Network inventory, the official media transparency documentation identifies two relevant limitations. Delivery reports can lag by one to two days, and contractual restrictions can exclude impression data from certain publishers. These are reasons to investigate coverage before treating every difference as a counting failure.
They are not explanations for every discrepancy. The useful distinction is between what the report says, what it leaves unknown, and what additional evidence would be needed to resolve the difference.
Three statements that should not be treated as equivalent
A publisher may be absent from a pre-campaign list, absent from a downloaded delivery report, or absent from actual delivery. Those statements describe different things.
The documented pre-campaign list contains the top one hundred publishers based on customizable historical impression data. It is a planning reference, not a complete inventory list or a prediction of exactly where a future campaign will run. Using it as a whitelist without configuring an actual restriction would confuse information with control.
A post-campaign report concerns delivery that has happened, but its disclosure limitations still matter. A missing publisher name cannot, by itself, prove that no impressions were delivered there. Equally, an unexplained gap cannot identify which undisclosed publisher received them.
For an international marketing team, that distinction affects the questions asked of an agency or platform representative. “Why was this absent from the historical list?” is a different question from “Can you confirm whether this placement received our ads?” A useful answer to the first does not settle the second.
Reconcile a finished period before judging the gap
Begin with the mundane checks: the same advertiser account, campaign selection, date range and time-zone interpretation. A broad account total and a filtered campaign export are not competing versions of the same number. Neither are a live dashboard and an earlier download.
Choose a completed period and allow for the documented reporting delay. Download that same period again rather than extending the range with another day of live activity. This makes it possible to see whether the original difference narrows as reporting catches up.
If it remains, preserve the difference. Do not distribute missing impressions proportionally across visible publishers just to produce a complete chart. That allocation would describe a model created by the analyst, not observed delivery. Without an explicit reason to model it, leaving a clearly labeled unknown is more informative.
The same caution applies to fraud conclusions. Missing disclosure and invalid traffic are separate issues. A report with incomplete placement detail may limit an investigation, but the gap is not evidence of a particular source of invalid activity. Ask the platform what coverage restrictions apply before making claims about where the missing volume went.
Decide whether the unknown prevents your intended use
The right response depends on the decision the report is supposed to support. A team checking a strict placement exclusion needs evidence about that restriction. If the available reporting cannot provide it, favorable campaign totals do not fill the gap. Review the supported exclusion settings and whether the remaining uncertainty is acceptable for that campaign.
A team comparing the disclosed major placements can still learn something, provided it states the scope. The difficulty arises when coverage changes between periods. An app's apparent share may rise because its delivery increased, because other delivery is no longer visible, or because both happened. Comparing percentages without checking the denominator can suggest a creative or budget problem that the data does not establish.
For B2B suppliers, there is another boundary: a placement report describes advertising exposure, not purchasing authority. Even a fully reconciled impression total would not tell you whether the audience included the right procurement roles. Keep that later business question separate from the immediate reporting question.
Our article on TikTok feed and cross-app placement environments discusses the environment distinction. Here the narrower task is to establish how much of a campaign's delivery a particular export can actually explain.
A concise discrepancy note is enough: the fixed period, export time, filters, observed gap and the platform's explanation of any disclosure limit. If the explanation is still pending, say so. The purpose of reconciliation is to understand the difference, not to make it disappear cosmetically.
Sources
TikTok for Business, About TikTok Ad Network media transparency solutions, documentation updated September 2026. The report's list size, usual delay and possible contractual exclusions are attributed to that source; no account-specific gap is estimated here.