Social Rocket
TikTok lets app advertisers specify, in eligible Smart+ campaigns, which measurement source counts for attribution and bid optimization. The Attribution Source Optimization feature, documented in October 2026, distinguishes between the source of the reported conversions and the event signals the algorithm uses to optimize delivery. According to TikTok Business Help Center, it is not yet available for all accounts.
Two decisions instead of one blanket conversion value
As the attribution source, a Mobile Measurement Partner (MMP) or TikTok's own SAN (Self-Attributing Network) reporting can be selected. As the data source, depending on the integration enabled, MMP signals, the TikTok App Events SDK or the Events API can be used. This separation matters when an app company evaluates its campaign results using AppsFlyer or Adjust, for example, while TikTok itself shows different conversion numbers.
The change is not an automatic alignment of all dashboards. Different attribution windows, identification options and privacy settings can still lead to diverging results. Its main effect is to make transparent which data TikTok should use to steer your campaigns.
Where the new switch actually works
The official guide names clear limits: App Promotion with an app install or retargeting objective is required, along with the updated Smart+ environment and a working conversion postback connection. The supported MMPs currently include AppsFlyer, Adjust and Singular; further partners are announced but not automatically integrated.
On Android, businesses can choose between MMP and SAN attribution under suitable conditions. On iOS, the choice depends more on whether only MMP signals or additionally SDK signals are available. For those who use Apple's SKAdNetwork as the authoritative reference, TikTok currently recommends the standard MMP configuration. For campaign-level budget control, the chosen sources must match across all ad groups.
What German app marketers should test before switching
First, paid social and analytics teams must define a binding reporting reference: who gets credit for an install or a purchase, and which time windows apply? Then postbacks, the iOS/Android split and conversion schemas need to be checked. A controlled comparison with an unchanged campaign makes sense; pure before-and-after differences can include seasonal and auction effects.
Important: Better agreement between reports does not yet prove additional revenue. Genuine advertising impact may require holdout or incrementality tests. This is also illustrated by the discussion of cross-platform retail media measurement at Lowe's and the changed Meta Ads attribution. For e-commerce on TikTok, separating creator sales, campaign touchpoints and shop tracking is also crucial.
Sources: The technical requirements and differences come from TikTok's product documentation and setup guide (both as of October 2026). A Digiday article sponsored by TikTok provided a research prompt, but it is not an independent effectiveness study.