Every Change You Make Should Solve a Problem

What is the problem to be solved?

I want this phrase to ring in your head every time you’re about to make a change. When you create a separate campaign. Create more than one ad set. Restrict by age or gender or placement. Use manual bidding. Or turn off that ad.

I want you to stop what you’re doing and ask yourself: “What problem does this solve?”

It’s not that there’s never a reason to do any of these things. The issue is that, far too often, advertisers do them based on feelings or vibes or assumptions.

Let me provide some examples.

I heard from someone the other day who was told to set up his campaign a specific way. One campaign with five ad sets, each with one ad. CBO was turned off.

When I asked him why he set it up that way, his answer leaned heavily into doing what he was told to do. It had something to do with creative testing and forcing Meta to show each ad.

My follow-up question was simple: “What problem does it solve?”

And the people I ask this tend to come up with a version of an explanation that is more focused on control than on an actual problem. And that control may be completely worthless in many cases.

If you’re optimizing for purchases, you’re splitting up your budget five ways to test ads. You will already struggle to get meaningful data, and now you’re all but guaranteeing it.

So you’re doing this to get low-quality data to act on which isn’t likely to impact your results in any positive way.

You’ve imagined a problem that may not exist and compounded that problem by creating a new one. The new problem is a combination of audience fragmentation, auction overlap, and a general watering down of your budget. All in the name of control.

Another example is restrictions by demographics.

I’ll often look at ad sets and see people restrict by age group when optimizing for purchases. When I ask why, I’ll usually get some form of explanation related to their ideal customer.

“People under 30 don’t buy” or “Our customers are almost always between 30 and 55.”

But what is the problem that they’re trying to solve? Did they run this ad set before and find that Meta wasted a large percentage of their budget on people under 30 or over 55?

When pressed, that’s almost never the answer I get. It’s a decision based on feelings.

You’re not even giving Meta a chance to find customers outside of your expected range. You might be surprised.

Make decisions like these based on data.

An example could be that you have data that shows that people under 30 have a much lower lifetime value. It’s not just what they bought today, but they buy less later. If this finding is based on meaningful data, then you have information that Meta doesn’t.

Because of that, Meta may spend budget on that group that could seem wasteful. You have a problem to be solved. But instead of restricting by age entirely, I’d use a value rule to bid less on that group instead.

A third example is removing placements.

When I see advertisers remove placements and I ask why, I often get one of a few possible answers.

“Our customers aren’t on Instagram.” Or “Audience Network is just bots.” Or “Facebook is a dead platform.”

But even if any of these things are true, it doesn’t necessarily reflect a problem to be solved.

Let’s assume you’re optimizing for some sort of a conversion. It really could be either a lead or a purchase.

If your customers aren’t on Instagram, Meta will limit your spend there. If Audience Network is just bots, and you might be right, Meta knows that you can’t get conversions there. If Facebook is a dead platform, Meta will know that you’re unlikely to find new customers there.

The performance goal matters a whole lot when assessing whether there’s a problem to be solved.

If you’re optimizing for ThruPlay Views, it matters if your customers aren’t on Instagram. Meta will show your ads there anyway because all that matters is that you get views.

If you’re optimizing for link clicks or landing page views, it matters if Audience Network is just bots. Meta will spend lots of money there to get you clicks.

But when you’re optimizing for conversions, there aren’t placements that are known to be sources for low-quality conversions. Especially if that conversion is a purchase.

Don’t remove placements based on feelings or vibes or assumptions. Only consider it because there’s a problem to be solved that can be tied directly to a placement. And even then, prioritize using value rules first before removing a placement entirely.

So here’s the bottom of the glass.

I want this phrase to become a daily part of your vocabulary. It doesn’t only need to apply to decisions made to your advertising.

But Meta advertisers are susceptible to unproductive work in the name of looking busy and useful. Don’t fall for it.

Before you add that restriction, ask “What problem does this solve?” And do we have meaningful data that proves it?

Before you add complexity by creating new campaigns or ad sets, ask yourself: “What problem does this solve?” Is it simply what’s expected? Or is there an actual reason based in performance?

Understand that there absolutely are reasons for restrictions and campaign complexity. But we shouldn’t apply any of them blindly.

Those who are most productive will apply them cautiously and only when making that change is a clear solution to a problem that needs fixing.