Campaign Optimization Using Non-Prioritized Facebook Events

One of the biggest Facebook advertiser complaints related to iOS 14 changes and Aggregated Event Measurement is the limitation to eight events per website for optimization purposes. Luckily, that is changing.

Let’s take a closer look at what the limitation meant, what happens with the change, and how this will impact your advertising.

The 8-Event Limit

When Aggregated Event Measurement was originally rolled out, advertisers were asked to provide a priority ranking of the eight pixel events they would use for ad set optimization.

Aggregated Event Measurement Ranking

Only these eight events would be allowed for optimization.

The ranking would be applied when collecting results from users who had opted out of iOS 14 tracking. In those cases, Facebook would only receive a user’s highest-ranking event, and receipt of that conversion would be delayed for up to 72 hours.

Problems Related to 8-Event Limit

Of course, there were many problems related to the 8-event limit. Those problems ranged from eight events not being enough to not being able to rank an event at all.

1. 8 Events Not Enough: Maybe you have dozens or hundreds of products. You may find success optimizing for the purchase of that specific product, rather than a standard Purchase event.

Or you may have a company that shares the same website across locations and departments. You don’t want to share the same eight events across those locations and departments.

2. Event Doesn’t Occur on Owned Website: This may be the most common. You have your own website, but the purchase itself occurs on a third-party domain that you don’t control. You can place your pixel on that website, but you can’t rank an event occurring there since you aren’t able to verify the domain (among other potential issues like sharing that event among other vendors).

Solution: Optimize for Non-Prioritized Events

Luckily, advertisers will now be allowed to optimize for events that are not included within their ranked eight (this feature is rolling out, and you may not have access to it yet).

There will be limitations to this approach that you’ll need to understand, but optimizing for non-prioritized events might be a great option for you.

Targeting Limitations for Non-Prioritzed Events

When optimizing for a non-prioritized event, your ad will not reach anyone on an iOS device who has opted out of tracking. That means that you WILL reach the following people:

  • Those on iOS device who opted in to tracking
  • Those on a non-iOS device

Keep in mind that this means you can also reach people who opted out of iOS tracking but are using a non-iOS device when your ad is shown.

The question is going to be how much this impacts your audience pool. If most of your traffic comes from non-iOS devices, this may be a great option. Even if you have a large iOS audience, it may be interesting to test.

Reporting Benefits of Non-Prioritized Events

The really interesting benefit of optimizing for non-prioritized events is that you won’t have to worry about reporting issues related to Aggregated Event Measurement.

As you know, AEM means reporting for users who opted out of tracking will be delayed and aggregated. It will take up to 72 hours to receive. You will only receive the highest ranking event. And other reporting may be modeled. This is a very inexact science. The uncertainty is frustrating.

But, what if you could get “normal,” pre-iOS 14 reporting? That’s essentially what happens when you optimize for a non-prioritized event. Reporting will be complete. Aggregated Event Measurement won’t apply.


Let’s see how this works. If you don’t have room in your ranked eight or simply have no ability to rank a given event, optimize for it as a non-prioritized event.

I’d take it a step further. Even if you don’t have these limitations, let’s try it anyway. Personally, I’m curious about the results we’ll see for non-prioritized event optimization vs. prioritized event optimization. Create a custom conversion, don’t rank it, and try to optimize for it.

Assuming you’re able to do this (not everyone has it so far), what results do you see?

Let me know in the comments below!