+1 vote
Unlike Facebook and Snap, Tiktok actually shows app install ads to Lat enabled users. However, when we export raw data from our MMP, we see that the percentage of installs attributed to Tiktok Ads is that don't have IDFA (that have been attributed through fingerprinting) is around 1%-2%. Our normal LAT on rate is around 70% (because our users are generally under 18 yo), so I'm confused on why the number we're seeing from TikTok is so different than the overall traffic. Do you think this is because Appsflyer is unable to fingerprint the installs accurately and they are attributed to "organic" or because only 1%-2% percent of the impressions of TikTok app install ads come from Lat Enabled users?
by (230 points)

1 Answer

+1 vote
Assuming TikTok is showing your ads to their LATon users, and that your campaign include those, the likely explanation here is that your MMP only uses IDFA matching to attribute, just like it happens on any SANs, without fingerprinting on top. So it's likely that users do see the ads, interact with them and install your app, you're just not seeing it. This is the same mechanic as the one in place for Apple Search Ads : LATon user are targeted, but not reported to MMP.

What's problematic isn't so much what you see/don't, but what TikTok does, for optimization: how is your discrepancy between Tiktok self-reported and your MMP?

If you don't want to run dubious guesstimates based on overall LAT rate, you may be able to extrapolate some of the invisible impact of TikTok Ads by isolating your organic-LATon cohort, and look at it side by side with LAToff organic & LAToff TikTok installs, and guesstimate the part you aren't seeing. Another signal, hardly comparable but useful for directional input, is to look at the numner of app units coming from TikTok in AppStore Connect Analytics (you'll see ads & organic there).

In any case, this is going to be irrelevant within a month or so.
by (2.4k points)
This makes perfect sense! Thank you for the suggestion.