Perceptycs clients running Meta app promotion campaigns
Measure Meta app ads with AEM instead of SKAN
Much more accurate; big accounts fully migrated
Workedanalytics-attributionpaid-social
What they did
Moved Meta app promotion campaigns from SKAdNetwork measurement to Meta's Aggregated Event Measurement (AEM). About 8-10 months before recording AEM was weak; 4-5 months before it was only recommended for smaller clients; by recording it was strong enough that some big accounts were fully migrated. Under SKAN, apps spending under ~10K/month missed privacy thresholds and postbacks.
What happened
Guest says AEM is now much more accurate than SKAN; several big accounts fully migrated.
In their words
Sub Club by RevenueCat · Making Web2App Work: From Basic Landing Pages to Complex Sales Funnels — Nathan Hudson, Perceptycs“there have been some big accounts that we fully migrated from SKAN to AEM, because it's that much more accurate.”
“Now it's really strong. So, yeah, whatever they've got going on, it works really well, and I encourage anyone who's pushing app promotion campaigns to try that on Meta.”
Related topics
Get tactics like this every Monday
New tactics from the week's founder interviews, each linked to where it was said.
More from this episode
- Advertise a free recorded masterclass, not the app, then sell the app at the endAgency says it worked, especially long-form on YouTube; no numbers shared
- Copy your working in-app onboarding to the web before inventing a new quizAgency rarely sees a profitable in-app flow fail when mirrored on web
- Launch your first paid ads to a web checkout so money and data come back fastPayouts in days, not ~60; about $1K of spend can teach you something
- Give each creator their own web page with tracking links instead of an App Store linkGuest calls influencers and affiliates 'amazing with web'; no numbers
- Run web sales ads and free guides to reach people who ignore app adsReaches new user pools and keeps data on non-buyers; no numbers
- Send sign-up, trial and purchase events from your web funnel back to MetaStandard practice; value-based bidding tried with no results yet
- Put screens between quiz questions that prove, reassure and show how the app works
- None to Run: Write a blog post that ranks for 'alternative to Couch to 5K' and links to the appBecame the app's main traffic source
- Elevate: Soft launch a rough version on Android and watch replays of every user~500 downloads a day; learned a lot without needing scale
- Charge $1 for the first month on web instead of a free trialScreens out fake cards; small net effect on conversion
- Dipsea: Test a money-back guarantee instead of a free trial on the webExperiment in progress; no results shared