Now that audience segments are available for manual Sales campaigns, it opens up all kinds of fun opportunities for testing and learning. That’s what this post is all about.
If you’re not familiar with audience segments, they’re set within your Ad Account Settings. You can define your Engaged Audience and Existing Customers.
This information can then be used to provide greater insight into your reporting. Using breakdowns, you can generate separate rows for each of these groups — as well as “New Audience” (those who qualify for neither group).
This transparency was helpful, even necessary, for Advantage+ Shopping Campaigns, which are driven almost entirely by algorithmic targeting without the advertiser’s input. These audience segments help answer questions about who is seeing our ads.
Now that audience segments are available for manual Sales campaigns, we can use this to answer some similar questions that have gone unanswered — until now.
Here are three ways you can leverage audience segments to get greater insights into your manual Sales campaign reporting. At the end, I’ll also provide some surprising results of what I’m seeing…
1. Advantage+ Audience without Suggestions
When you create an ad set that uses Advantage+ Audience, you have the option of providing an audience suggestion.
If you don’t, ad delivery will be entirely algorithmic. Meta says that their “AI uses lots of information to find your audience” — like past conversions, pixel data, and prior engagement with your ads.
Here’s a screenshot of that explanation…
That sounds a whole lot like remarketing, right? In other words, even if you don’t provide an audience suggestion, Meta’s AI should — in theory — prioritize showing ads to people you’d otherwise select to target.
I’ve long wondered whether using an audience suggestion mattered. I’ve decided that while it may not make a difference, it can’t hurt.
But, what actually happens? Does Meta’s AI prioritize remarketing audiences like their documentation claims?
Thanks to audience segments, we can test this. Define your audience segments as thoroughly as possible.
This is how I defined my Engaged Audience…
And my Existing Customers…
Next, create a Sales campaign with Advantage+ Audience without providing an audience suggestion. You will then be able to use Breakdowns by Audience Segment to see how many of the people you reached fall within Engaged Audience, Existing Customers, or New Audience.
2. Advantage+ Audience with Suggestions
We can also use audience segments to help answer our questions about whether providing audience suggestions makes a difference.
As I said in the prior section, I tend to use audience suggestions. It’s partly out of habit. But it’s also out of a belief that, at best, it can help the algorithm get started. At worst, it shouldn’t hurt.
Back to Meta’s documentation. If you provide an audience suggestion, Meta says that they will “prioritize audiences matching your suggestions, before searching more broadly.”
Again, let’s screenshot this for emphasis…
In theory, if we were to provide suggestions matching our Engaged Audience and Existing Customers, we should see Meta’s explanation above reflected in our breakdown by audience segments.
So, let’s do that! Create a Sales campaign using Advantage+ Audience. Provide audience suggestions that match your definitions of Engaged Audience and Existing Customers exactly.
The reason for this approach is simple. There’s no reason to provide detailed targeting or lookalike audiences as suggestions since we can’t use those to define our audience segments. Because of that, we’ll never know for sure whether people in those audiences saw our ads.
Since we’re told that Meta AI will prioritize our audience suggestions before going broader, we can prove that one way or another by using the exact custom audiences for suggestions that we used to define our audience segments. When we breakdown our results, we should see that reflected.
In theory, of course.
3. Original Audiences with Advantage Custom Audience
I’ve mostly abandoned original audiences (and Advantage expansion tools that go with them) since the rollout of Advantage+ Audience.
Original audiences feel like old strategies, and we should use Meta’s new and improved tools. Advantage+ Audience works in much the same way that Advantage Detailed Targeting, Advantage Lookalike, and Advantage Custom Audience, but Meta says that Advantage+ Audience is better and more advanced.
Back to Meta’s own documentation on Advantage+ Audience, this is spelled out:
Meta’s original audience options, including Advantage options (Advantage detailed targeting, Advantage custom audience and Advantage lookalike), can limit the potential of Meta’s AI which can be less effective.
Based on Meta’s own words, we assume that these work similarly, but Advantage+ Audience has the ability to go much broader (and lead to better results). So, the assumption is that if you turn on audience expansion with original audiences, the audience will expand — but your original inputs may be more respected.
Once again, we need to stick with the topic of custom audiences since these are what can be verified with audience segments. If we provide all of the same custom audiences that were used in our audience segments and turn on Advantage Custom Audience, what would happen?
How many of the people reached would be from our custom audiences? How many would be from expansion? And how does this compare to when using Advantage+ Audience?
We can test this! Once this is set up, use the breakdown by audience segment to see how your ads are distributed.
Initial Learning
I actually started part of this test already. The early results represent a small sample size, and in some cases they have been surprising.
It’s not clear how much the conversion event matters. Will your Engaged Audience and Existing Customers be used differently depending on whether you optimize for a purchase, lead, or something else?
Other factors like the sizes of the audience segments, sizes of the custom audiences used for suggestions, budget, and time may all contribute.
My initial test used a custom event for 60 second website views as the conversion event. The results were staggering. When providing audience suggestions, less than 1% of my budget was spent on them. When providing no suggestions, it was only slightly better.
But, I started a new test and the results have (thankfully) adjusted. Distribution to my Engaged Audience and Existing Customers has increased significantly, regardless of which approach I’m taking. These results have increased my faith in Meta’s claims that remarketing happens, regardless of whether you provide audience suggestions.
I’ll hold off on sharing specifics until I’m done. Until then, I encourage you to test this, too.
A Note on “Sales” Campaigns
Something that flies a bit below the radar is that you don’t technically need to optimize for purchases when running a Sales campaign. Because of that, you could run tests like I describe in this post while optimizing for any website conversion event (leads, registrations, custom events, and whatever else you use).
Sales is simply how you defined your campaign objective.
It doesn’t determine how your ads are optimized. This is defined by your performance goal and conversion event.
This is the case with Advantage+ Shopping Campaigns, too. Yes, it’s super confusing. You don’t need to optimize for SALES when running Advantage+ Shopping or manual Sales campaigns.
Your Turn
Have you run a test like this? What have you seen?
Let me know in the comments below!
