Meta Ads Targeting: How to Approach Them : Ankita Mankotia

Meta advertising is nothing like it once was. The role of the advertiser has changed. New tools and features have emerged. The strategies have evolved. And if you approach meta ads targeting now like it’s 2018 or 2020, you’re going to struggle.

Unfortunately, that’s exactly what we see. Some advertisers have embraced this brave new world. Others are resistant to it and insist on forcing their old strategies like a square peg in a round hole.

It’s time to wake up. In this post, we’ll evaluate how meta ads targeting has changed and how you should approach it now.

Meta Ads Targeting Before

Back in the day, there were three distinct buckets of Meta Ads Targeting.

1. Cold Targeting

We loved uncovering the magical combination of interests, behaviors, and lookalike audiences to bring the best results from a cold audience. This is the group of people who can potentially see your ads.

You help influence this by adjusting age, gender, location, detailed targeting (interests and behaviors), custom audiences, and more.

We experimented with grouping interests in one ad set and lookalikes in another. Or we’d layer interests on top of lookalikes. Should you use a 1% lookalike or 5%? What about 10%? We tested and tested and found the answer.

Even location, age, and gender were important details. A part of the country isn’t leading to conversions. Examples of conversions include purchases, leads, content views, add-to-cart, and registrations. Exclude it. Mainly women between 25 and 34 are buying. We’ll only target them.

2. Warm Targeting

If you wanted to go after a group of people who knew who you were and were likely to convert, there were several places to start. Target your page followers, anyone who engaged with your page or posts, people on your email list, or anyone who visited your website.

This was a go-to meta ads targeting strategy.

3. Hot Targeting

These people are hot for a reason. They performed a very specific action. I’d push people through several stages of a campaign, showing them a different ad every few days. And it worked great!

There was a good reason to use all three approaches. It was generally seen as good practice to have multiple ad sets, if not multiple campaigns, dedicated to each audience segment.

The Evolution of Meta Ads Targeting

There were a couple of turning points. One was the Cambridge Analytica scandal. While it happened in and around 2015, it wasn’t revealed until 2018 and the impact to targeting would come after. One of the main lessons was to prevent bad actors from using sensitive targeting to manipulate elections.

Another turning point was iOS 14 and the movement towards greater online privacy generally. Facebook would face greater scrutiny regarding what was collected, how it was used, and giving users more control.

These combined forces led, directly or indirectly, to the removal of thousands of interests and the inability to target specific groups when a special ad category is involved. Opt-outs also cut into remarketing audiences, making them less complete and less dependable.

In the meantime, Facebook — and eventually Meta — would need to come up with solutions that would overcome these disadvantages. That led to a focus on AI, machine learning, and expanded audiences.

The move towards broad targeting began with Advantage Detailed Targeting, Advantage Lookalike, and Advantage Custom Audience. You provided targeting, but the algorithm would be able to reach people beyond that group if it would lead to more results.

The next step was Advantage+ Shopping Campaigns, which virtually eliminated targeting inputs completely. Beyond having some say over how much you’d reach current customers, the algorithm had entire countries of users to target without restrictions.

Eventually, this same approach would begin rolling out to any campaign objective in the form of Advantage+ Audience. You can provide targeting suggestions, but otherwise the algorithm will use pixel data, conversion history, and ad engagement history to build a starting audience.

How to Approach Cold Targeting

An argument can certainly be made that there’s very little reason for interests and lookalike audiences now. But even if you use them, there’s no reason to use them the way we did before.

You don’t need to obsess over which interest is most effective because, in most cases, Advantage Detailed Targeting is automatically on and can expand your audience anyway.

There’s no reason to constantly test different lookalike audiences and percentages because Advantage Lookalike is often on by default, which will expand the percentage if necessary.

The evolution towards broad and expanded audiences changes our approach, whether you like it or not.

There’s simply no reason to spend much time on testing different audiences since the algorithm can go beyond the audience you use anyway.

It’s a complete waste of time to have multiple ad sets for different cold targeting approaches when the overlap is likely to be significant and audience fragmentation may result.

What should you do?

Embrace broad targeting for cold audiences. If you’re optimizing for a purchase, test Advantage+ Shopping Campaigns.

Otherwise, use Advantage+ Audience. Add some targeting suggestions if you want. But the true power will be how the algorithm learns beyond that initial group.

I wouldn’t be surprised if we eventually see the elimination of Advantage Detailed Targeting and Advantage Lookalike in favor of Advantage+ Targeting only since the functionality is similar and confusing.

But otherwise, you should embrace the expansion of your audiences when given the option.

Bottom line…

1. Create fewer ad sets for the purpose of cold audience segmentation.

2. Embrace expanded audiences when given the option for cold targeting.

3. Embrace machine learning and AI for the broadest of targeting.

Is Remarketing Dead?

This is a common refrain, and it’s at least partially valid.

Generally remarketing is mostly unnecessary. What I mean by that is that it probably isn’t necessary to target the “warm” audiences we defined at the top of this post. These are the types of groups that will be built into the initial focus of broad targeting.

You could make an argument to use some of these remarketing audiences in testing. For example, target all website visitors and turn on Advantage Custom Audience.

Or provide a group of custom audiences as your targeting suggestions when using Advantage+ Audience. In both cases, though, it’s a matter of using this group as a starting point with the hope that it helps the algorithm.

We’ll figure out with time whether using custom audiences in these ways was beneficial or if the algorithm would have searched the most valuable people in those groups out anyway. But for now, it doesn’t hurt to experiment with this.

Something I haven’t bought into is abandoning remarketing completely. I still subscribe to abandoned cart remarketing for simple reasons: It works, it’s inexpensive, and it’s very profitable.

If you have a small budget, the broad targeting approach isn’t likely to yield many conversions. But you can spend a very limited amount by retargeting people who abandoned the cart and get results.

Maybe I’ll change my stance on this eventually. For now, I still see remarketing to the hottest of audiences makes a ton of sense.

Your Thoughts

How has your meta ads targeting approach evolved?

Let me know in the comments below!