+2 votes
My question is regarding LAT users and how you calculate these in your reporting across different channels. Obviously with ASA it is nice and easy due to their reporting. But how about for FB, UAC, Snap, TikTok, etc? If a LAT user is aquired through these channels are there any extra attribution methods in play (fingerprinting etc), especially as an MMP for FB?

Using FB as an example, will FB be showing users with LAT On, specific targeting ads via lookalikes? If not then using a blanket 10-30% LAT average multiplier (dependent on your app) wouldn't be accurate.
by (240 points)

1 Answer

+2 votes

Try by yourself to apply a multiplier to all your UA, and your organic may become negative
There are 3 groups of channels which behaves very differently in this regard:
- ASA has a very unique way of dealing with LAT: show ads, but don't report. I don't think is "nice and easy" at all in this regard, but if you're happy extrapolating from Apple numbers, go for it
- Networks where app installs ads are (mostly*) not shown to LAT users, which include: Facebook, UAC (display & youtube), programmatic & remarketing networks. If you extrapolate anything here, you're making it up. If you only advertise there, you're missing out on a large chunk of your audience
- SDK-adwords, anything web, content marketing, others: ads are shown to LAT users, and captured via fingerprint & similar techniques. While it's far from perfect (and SANs are far from it too), it somewhat works and you do see LAT attributed there, so I wouldn't extrapolate either. Note that Facebook for web would be in this group, the above one is only for App Installs.


Notes:
* there are always discpreancies and it's usual to see a small 1-2% LAT rates on those (vs 20-30% in the market)
- This is based on my observations, I'm happy to be proven wrong on specifics so I can keep learning
- I tried to go through those in a recent interview with Shamanth on the MobileAcquisitionShow)
- I'm talking for iOS here, given that Android rates are now very low (<3% according to the latest Singular report) and not an area of focus for UA for the apps I work with 

- UAC Search would be a fourth group, but the pattern was really bizarre and this isn't reported to MMP since January so irrelevant for the future

by (2k points)
Amazing answer and I do appreciate your time to do this!! ;)