Targeting Isn't What It Used to Be

So yeah, let's talk about targeting.

This is a topic I've covered in bits and pieces in prior episodes, but I really want to take a broader look at it. Today, I see far too many advertisers who obsess over targeting, who think it's far more important than it is, and who make their results worse in the process.

Honestly, all you're doing is stressing yourself out unnecessarily. You're making it more complicated than it needs to be, just for the sake of feeling busy.

The reality is this: If you're still approaching targeting like it's 2017, we've got a problem.

Back then, targeting was the strategy. I admit it. It was central to everything I did. There are still social media quotes of mine floating around saying that targeting is the most important part of your ad's success. And that was true.

We'd load up on relevant interests and behaviors. We'd split test 50 different lookalikes. And we felt like geniuses when it worked.

But now, that strategy is mostly window dressing.

If you're optimizing for conversions, most of your targeting inputs are treated as suggestions. Suggestions, not rules. So you plug in a lookalike, and Meta says, "Cool," and then expands far beyond it if that will help you reach your goal. Same with detailed targeting.

Whether you provide detailed targeting inputs or not, your ads are likely reaching the same people. And you know what that means? There's rarely a point in segmenting targeting into separate ad sets anymore. If the algorithm can reach the same people across ad sets, all you're doing is creating overlap.

Now your budget is split up. That won't help you exit the learning phase. And now your ads are competing with one another. It's not strategic. It's wasteful.

Let’s talk about age and gender.

Are you still restricting by age because you think your ideal customer is 35 to 44? Still excluding men because women convert better? You probably don't need to.

If you're optimizing for sales, the algorithm will figure it out. You've already trained it with conversion data. It knows who's buying.

Restricting age and gender can backfire. It creates limitations. It adds friction. And it often increases your costs.

The one time when this can get messy is if you're optimizing for leads.

When you do that, Meta focuses on quantity. Its goal is to get you the cheapest leads possible. But if those leads are complete garbage — wrong demo, no intent, low quality — you may need to step in.

That might mean restricting age or gender. It might mean optimizing for a deeper event. But only make those changes when there's a clear problem. Otherwise, stay hands off.

Here’s the targeting setup I recommend:

Start with location.
If you're national, go broad. Include the full countries. Don’t get cute by isolating the states or regions you think will get you the best results.

If you're local, target those locations. Just know that local business targeting isn't perfect. It includes people who live there and people just passing through.

Age and gender? Leave it wide open.
Unless you've proven otherwise with testing and confirmed that it improves performance, don’t touch it.

Detailed targeting and lookalikes?
Sure, add them as suggestions if you want. Knock yourself out. But don’t expect them to do much. They’re not what they used to be.

Please don’t segment by ad set unless it’s absolutely necessary.
Don’t split up interest groups or lookalikes into their own ad sets. There’s no point. Meta is going to expand anyway.

Don’t run one ad set for men and one for women. Same with age. Unless there's a very good reason, all you're doing is watering down your budget and making your ads compete against each other.

If you want better targeting, put it in your ad creative.

That’s where it happens. Speak directly to the people you want. Repel the ones you don’t. Make sure the right people complete the action you want them to take. That comes down to your ad copy, your creative, and your offer.

That’s where targeting lives now.

So this is the bottom of the glass.

Detailed targeting and lookalikes don’t do much when you’re optimizing for conversions. Meta treats them as suggestions and fills in the gaps on its own.

If you’re segmenting audiences, splitting up ad sets, narrowing age and gender, you’re probably not helping. You’re just giving the system more to work around.

If your results aren’t broken, stop fixing them. Let the algorithm breathe. Let it learn.

If you want to control who sees your ads, speak to the people in your creative. That’s where the real targeting happens now.

If you want to learn more about this and how to approach targeting today, check out my recent blog post at jonloomer.com/targeting.