Meta has begun to roll out an update to ad sets utilizing performance goals for link clicks and landing page views, which was originally announced in January. When using original audiences, Advantage Detailed Targeting is automatically applied.
If you missed that Advantage Detailed Targeting was turned on, you’d be forgiven. This design variation is not at all obvious. If you miss the new label, you won’t see that the audience may expand unless you hover over one of the tooltips.
This is a departure from the primary design I’ve seen when using conversion performance goals. In those cases, a message is highlighted in gray.
But, let’s back up. There’s plenty to unpack here. The signs are all around us that we’re headed towards a future of less targeting control, regardless of the performance goal. And that could be a problem, unless Meta makes some much-needed improvements.
In this post, let’s discuss:
- The current state of audience expansion
- Where expansion is effective
- Where expansion fails
- Where this is headed
- What Meta needs to do
At the bottom of this post, I’ve also recorded a video that summarizes what is going on.
Current State of Audience Expansion
Meta first unveiled audience expansion in 2021 with a suite of products that would eventually fall under the “Advantage” line. Here is how they work…
Advantage Detailed Targeting
Advertisers provide detailed targeting inputs that Meta prioritizes. Your audience can be expanded to reach people beyond that group if better results can be found.
Not long after its initial rollout, Advantage Detailed Targeting became a fixed default for any conversion-related performance goal. Otherwise, advertisers had the option of turning it on or off.
That, of course, changed with this latest rollout of Advantage Detailed Targeting for ad sets utilizing link clicks or landing page views performance goals.
Advantage Lookalike
The second of the Advantage expansion family, Advantage Lookalike works in a similar manner as Advantage Detailed Targeting. If Meta detects that better results can be found beyond the selected percentage of your lookalike audience, the percentage can be expanded.
For example, if you use a 1% lookalike, the audience could be expanded to anywhere from 2 to 10%.
Like Advantage Detailed Targeting, Advantage Lookalike is turned on by default for conversion performance goals and cannot be turned off.
The latest update to link click and landing page view performance goals has not been applied to lookalike audiences. Advertisers still have the option of turning this on or off in that case (for now).
Advantage Custom Audience
Next, Meta rolled out the ability to expand custom audiences if better results can be found. Unlike the first two features, there is always an option to turn Advantage Custom Audience on or off. There isn’t currently a case where it’s on by default (though this may change).
Of course, Meta then took things even further…
Advantage+ Audience
Beginning in August of 2023, Advantage+ Audience became the default way of selecting an audience in the ad set. Advertisers still have the ability to switch back to original audiences, where the three Advantage expansion tools may be applied.
When using Advantage+ Audience, any targeting inputs provided are seen as mere suggestions. You will reach people beyond that initial group, and providing suggestions is optional. If you don’t provide them, Meta will automatically begin with your pixel data, conversion history, and prior engagement with ads as a guide.
Advantage+ Audience is the initial default for all campaign objectives, regardless of the performance goal. When used, any targeting inputs — custom audiences, lookalike audiences, detailed targeting, and even gender and age maximum — are seen as audience suggestions, and your ads may reach people beyond those groups.
Where Expansion is Effective
While I initially resisted audience expansion (“I only want to reach the people I’m targeting!”), I’ve come around to it. But, it’s most effective for a unique set of circumstances.
Audience expansion (any of the Advantage expansion tools or Advantage+ Audience) can work because the algorithm is hyper-focused on finding your desired action, as defined by the performance goal. Your targeting constraints could conceivably restrict the algorithm from getting more of those actions.
This is especially true when optimizing for purchases.
Success is defined by getting more purchases within your budget. If your targeting can be expanded to find more purchases, that’s a good thing.
There’s no better example of this in action than Advantage+ Shopping Campaigns. Targeting inputs are virtually nonexistent, and yet Meta says that they lead to a 17% improvement in cost per acquisition and a 32% increase in return on ad spend.
Where Expansion Fails
For the same reason that audience expansion can be effective for purchase optimization, it often fails for anything else — especially when using a performance goal that represents a top-of-the-funnel action (link clicks, landing page views, ThruPlay, post engagement, and more).
The audience will expand beyond your inputs if more of the actions defined by the performance goal can be found.
This isn’t a problem when optimizing for purchases because getting the purchase is the ultimate determinant of success. The algorithm makes adjustments based on whether it can get you more purchases.
It’s a problem for everything else because quality then matters…
Your audience is expanded to get more link clicks or landing page views. But did these people do anything else after clicking? Were they bots? Where they accidental clicks? Were they people who click everything? The algorithm doesn’t care.
Your audience is expanded to get more people to engage with your post. But is this positive or negative engagement? Do they fit your typical customer profile? Is there any chance that they’d ever buy from you? The algorithm doesn’t care.
Your audience is expanded to get more leads. But were the email addresses provided valid? Are these people reachable? Will they open their messages and engage? Is there any chance they’d ever buy from you? The algorithm doesn’t care (unless you optimize for conversion leads, which isn’t reasonable for everyone).
In each case, you care. And that’s the problem. Audience expansion fails when there’s no control for quality. Your targeting inputs were the only remaining constraints to focus only on potential customers.
Where This is Headed
Look to the most recent developments to predict where this is heading…
1. Advantage+ Shopping Campaigns don’t allow for any targeting inputs.
2. Advantage+ Audience is now the default, and you’re discouraged from switching to original audiences.
3. Advantage Detailed Targeting is now on and can’t be turned off when using link click and landing page view performance goals.
Every new update puts less importance on your targeting inputs. More ways to expand the audience. Fewer controls to be able to target an exact group.
Given that Advantage+ Audience is the default for all objectives and performance goals now, I’m actually surprised that Meta would make this update to Advantage Detailed Targeting related to link clicks and landing page views.
My assumption is that the ability to switch back to original audiences (and utilize Advantage Detailed Targeting, Advantage Lookalike, and Advantage Custom Audience) will eventually be eliminated. But, maybe this is a sign that such a move is further off in the future than I expect.
The bottom line is that Meta isn’t going to stop expanding audiences beyond your targeting inputs any time soon. We’re likely to see this forced for more objectives and performance goals in the future, even if you’ll be able to continue switching back to original audiences.
What Meta Needs to Do
I am not a fan of this latest update to Advantage Detailed Targeting. The reason can be found within the section about when audience expansion fails.
Optimizing for top-of-funnel actions is already problematic. But if Meta removes or de-emphasizes targeting constraints, we lose all checks on quality. It no longer matters who these people are. Meta only cares that they’ll perform the action that we want.
The solution isn’t that complicated, and it’s been needed for years. The evolution of audience expansion only makes it more imperative that Meta act on it.
There must be a way to optimize for quality top-of-funnel actions.
I’d be much more willing to use the link click or landing page view performance goals to promote my blog posts if I could require the algorithm to optimize for quality traffic — not just any traffic. This could be defined by time spent on the website, scroll depth, other conversions, and return visits.
I’d be much more willing to use performance goals related to post engagement if I could require the algorithm to optimize for quality engagement — not just any engagement. I want people who are likely to share my posts, provide thoughtful comments, and return to my content later.
This “quality” element could be an option when setting a performance goal. Do you care more about getting a high volume of actions? Leave it at the default. Do you care about quality? Check this box and expect to spend more.
If that were possible, the expansion of your audience becomes less problematic. The algorithm would expand to get more of the quality actions that you are wanting — and that is ultimately what would guide ad delivery.
This would seem like a natural solution that is good for everyone. Most importantly, advertisers would be willing to spend much more on actions other than conversions if there were an increased confidence in the quality.
Watch Video
I recorded a video about this, too, and you can watch it below…
Your Turn
Do you run ad sets optimized for link clicks and landing page views? What do you think about this update?
Let me know in the comments below!