A good case with AEM in Meta Ads

Last week we had a good case with AEM in Meta Ads (check my last posts to see the whole story!) so I applied the strategy with a different app but in this case… the results were not successful.

I am working with a client who recently launched an app. This means that they have NO presence at all in UK, which is the country that we are targeting in this test.

This is VERY evident if you go to Revenue Cat and check the data (see screenshot). As you can see, we had zero activity in the country until October 16th, when we decided to launch the first campaign in Meta using AEM and focusing on free trials.

If you count the trials from 16th to 20th, you will see 90 free trials in RevenueCat while if you go to AEM, I just have 33 trials in my campaign. This huge difference is not the worst thing, the worst is that SKAN reports 52 trials with an 11% null rate. So summarizing:

👉 Meta AEM reports 63% fewer conversions than RevenueCat and 34% less than SKAN, causing a worse performance due to a lower amount of events per day
👉SKAN only reports 42% less than RevenueCat, which looks like a more reliable attribution method when you optimize for low-funnel events

But for me, the main question of all of this is:

How is it possible that Meta reports less events than SKAN if apparently AEM uses, among other sources, SKAN signals to enrich the modelling? It would have to report at least the same amount of events than SKAN but it is not the case.

Things get more complicated because I have the same campaign for US and in this case, my AEM campaign is doing the correct job! It is reporting a bit more than SKAN which is expected since I am using AEM technology.

So another big question is:

Is this issue only happening when you run AEM campaigns in UK?

Let me know your thoughts and if you had similar experiences!

Have a nice week!