Common Sense Playbook to Meta Advertising

My approach to Meta advertising and the philosophies that guide me aim to provide clarity for something that is far too often confused. If you read my blog posts, watch my videos, or listen to my podcast, there are common themes of simplicity and responsibility.

Practice a simplified approach in favor of overcomplicating everything.

Understand how the algorithm works and leverage it, rather than hoping for it to work a different way.

Take responsibility for the things that you control and let go of those you don’t.

My hope is that when you approach Meta advertising in this way, it will make a whole lot more sense. This basic understanding will guide how you create campaigns, what you customize, and how you evaluate results.

If you’re struggling with finding direction in your Meta advertising approach, this post was written for you. Consider it your Common Sense Playbook to Meta Advertising…

The Algorithm is Literal

This is foundational to everything here. Understanding it directs what we do and don’t do. It’s almost too simplistic, but this phrase guides my understanding of Meta advertising.

“The algorithm is literal” means that the way your ads are delivered is easily explained. You define what you want with a performance goal. The algorithm will then do all it can to get as many of that action as it can within your budget.

Meta says exactly what the algorithm plans to do. It’s going to show your ads to people who are most likely to perform the action that you want. So define that goal carefully.

This can be both good and bad. It’s great when your performance goal is to Maximize Number (or Value) of Conversions and the conversion is a purchase. The algorithm is going to do what it can to find you more purchases.

But it can also be bad when optimizing for almost anything else. And it can explain why things didn’t go the way you expected.

Why am I getting low-quality leads? Because the algorithm doesn’t care if you get high-quality leads.

Why are my ads being shown primarily to an older demographic? Because the algorithm can find cheap actions from them, as defined by your performance goal.

Why is my budget getting wasted on the Audience Network placement? Because you defined Link Clicks as your goal and this placement can get you lots of them.

When looking for an explanation for your results, it’s easy to find when using this lens. The algorithm is doing what you told it to do. That could mean taking advantage of weaknesses to get you more results.

Once this is burned into you, your entire strategy will be far more intentional. There will also be far fewer surprises.

Read and Listen

I covered this extensively on a recent podcast episode. It’s a short listen, and there’s also a written transcript.

Listen here: Understanding How Meta’s Algorithm Really Works

Your Targeting Obsession is Rarely Necessary

Advertisers have a common urge to manipulate targeting. They want to very clearly define who will see their ads. The assumption is that this is a critical step to getting good results.

At one time, this was certainly true. It was central to my strategy. I often said that the perfect ad can’t overcome bad targeting.

But things are different now. While the algorithm is better at finding our ideal audience, the most important change isn’t the quality of that evolution. It’s that in most cases, your targeting inputs simply don’t matter that much.

If you turn on Advantage+ Audience, there are only three elements within your control:

• Location
• Excluded custom audiences
• Age minimum (no higher than 25)

Every other input you provide is an audience suggestion.

Gender, age maximum, custom audiences, lookalike audiences, and detailed targeting. Meta can look at and ignore them. In many cases, I’ve found that’s exactly what happens.

Oh! you say. But you can also turn Advantage+ Audience off. This can give you more control over age ranges, gender, and custom audiences, which will be respected. But depending on your performance goal, the detailed targeting and lookalike audiences may be purely decoration.

This is the direction we’re heading. If you’re expecting that we’ll eventually regain control over targeting, you’re living in a fantasy world. Algorithmic targeting is the present, and it will become impossible to avoid in the future.

Embrace this simple fact: Your targeting inputs just aren’t that important.

You need to know when targeting inputs are respected and when they’re not. You also need to know whether your inputs are a productive restriction or if they’re preventing you from getting better results.

If your entire strategy is built around remarketing, like mine once was, it shows a basic misunderstanding how how things work now. You don’t need to target general remarketing audiences (all of your website visitors or email list, for example) because the algorithm does that naturally. That doesn’t mean all remarketing should be eliminated from your toolbox, but it should collect far more dust than it once did.

You shouldn’t be surprised to learn that when your targeting inputs are the most helpful and when they’re detrimental has everything to do with that familiar phrase: The algorithm is literal.

Because the algorithm is literal, you want to limit the restrictions of targeting inputs when optimizing for a purchase.

Because the algorithm is literal, targeting restrictions may be required when optimizing for something else.

The main thing is that you should be consistent. Know why those inputs are or are not necessary, and base your approach on results rather than a gut feel. And prioritize limiting those restrictions when you can.

Read More

My outlook on targeting has completely transformed during the past year or two, and that’s reflected in the volume of content that I’ve published on it. I can’t isolate one piece of content that explains this best, so I’ve linked to a few below:

Attribution is Your Responsibility

The algorithm is literal. You define the action that you want. You also need to take responsibility for putting the structure in place to optimize for and report on those results.

Attribution is how Meta gives credit to ads for a conversion. Meta looks for conversions based on how you’ve defined them. Results are reliant on you defining them thoroughly and accurately. Changes are made to delivery when results aren’t coming, which is based on your work.

Meta doesn’t magically know that a purchase was made after someone was shown your ad. You put in the work on the back end to make sure Meta is notified when the purchase happens.

That requires doing at least the bare minimum of setting up the pixel on your website. You then need to define every important action with standard and custom events to be sure Meta is notified when these actions happen. And because the pixel alone will have holes, you will want to supplement what Meta has with first-party data using the Conversions API.

It’s amazing what a difference this can make. The algorithm is on a quest to do what you asked it to do. It requires that this be set up correctly. If not, it cannot report on every conversion that happens. The algorithm would then assume changes need to be made to who sees your ads. It will keep adjusting in an attempt to make you happy.

If you fail at properly and thoroughly setting up everything that is required for attribution, it won’t matter how you structure your campaign. No targeting or split testing of ad copy and creative will be able to overcome it.

And when you fail to get results, you’ll likely do what’s common: You’ll blame Meta.

Don’t let this be you. Make sure you’ve set this up properly.

Read More

I’ve written two guides that should be your starting point for conversion attribution:

Results Don’t Lie, But Advertisers Do

A common complaint I hear from advertisers is that results are wrong. But the results reflect reality, just not the reality that you want it to reflect. You have a job to provide context.

Meta doesn’t underreport results. You may not understand the limitations of reporting windows as defined by the attribution setting. You may not be sending first-party data that will help fill in the blanks. But Meta isn’t intentionally suppressing results.

And Meta’s reputation for displaying inaccurate results can also go in the opposite direction…

Meta doesn’t inflate results. You may have set up your conversion events wrong. You may have failed to properly deduplicate those events. Or you have failed to break down your results by view-through, click-through, number of days, and first vs. all conversions to make better sense of them.

The problem isn’t Meta doing either of these things. The problem is that advertisers, either knowingly or ignorantly, fail to execute their responsibilities when it comes to conversion reporting.

Incompetent advertisers absolutely will ignore the reasons behind results that seem too good to be true, hiding them from clients in an effort to make themselves look good.

And advertisers absolutely will blame Meta when results are terrible, ignoring the responsibilities that they neglected.

Read More

You can’t take conversion results at face value. The results may be accurate, but advertisers can still manipulate them to hide the truth.

How to Uncover the Truth Behind Conversion Results

Prioritize a Simplified Approach

A simplified approach is central to the Common Sense Playbook to Meta Advertising.

If your goal is to generate more purchases, your job should be straight-forward. There’s no reason to get cute with your performance goal, targeting, or placements. And since multiple ad sets in such a setup will lead to Auction Overlap, you can usually get away with a single ad set.

That doesn’t mean that getting good results is easy. But it does help focus your responsibilities.

Simplify the number of campaigns and ad sets. Consolidate your budget as much as possible to give whatever you’re running the best chance for success.

Focus on campaigns that optimize for some kind of conversion. Only run ads that are optimized for top-of-funnel actions (awareness, traffic, and engagement) if they aren’t stealing budget from your most important ad sets.

The advertiser’s basic instinct is that this can’t be right. Our constant tinkering and micromanagement are required. That unnecessarily complex campaign construction is needed to prove to your client that you’re doing something important. If we’re not doing all of these things, why are we needed at all?

But your services continue to be valued. You can then dedicate your time doing things that make the most impact.

You can create ads that inspire the action that you want from your ideal customer. Develop an offer that can’t be refused. Design a high-performing landing page that minimizes distractions and gets results. And get everything associated with attribution right.

Read More

A simplified approach makes a whole lot more sense once you understand how the algorithm works and your place in it. Read this post for a full breakdown:

A Simplified Meta Ads Strategy for Optimal Results

Have a Plan for When Changes Are Necessary

If you’ve read this far, you should understand that my recommendation isn’t to take a universally hands-off approach. Sometimes you should trust the algorithm and let it do what it does best. Other times, you should understand its weaknesses, and help guide it to where you need it to go.

The main thing is that we aren’t doing this randomly. We don’t want chaos. There are very clear reasons for when we’re hands-off and when we customize that are baked into how the algorithm works.

Audience Network is prone to driving low-quality traffic, but that doesn’t mean that you should always remove it. If you understand the strengths and weaknesses of the algorithm, you understand why.

Your business predominantly serves women, but that doesn’t mean that you should always restrict your audience by gender. When you understand how the algorithm works, you’ll know when you should and shouldn’t make these adjustments.

I would typically turn on Advantage+ Audience and let the algorithm do the heavy lifting, but there are exceptions. Since, depending on the optimization, the algorithm may not care about quality actions, there are times when restrictions are needed.

Once you understand what is outlined in this post, all of this will make more sense. You’ll have a much clearer and consistent approach that is based on how things actually work, rather than a gut distrust of an evil algorithm.

Read More

This post lays out many of the specific exceptions to when you should trust the algorithm and break away from Meta’s own best practices:

When to Break from Meta Advertising Defaults: A Guide

Execute the Playbook

I don’t expect this approach to be particularly popular because it goes against the grain of how advertisers have historically approached things. While that complex and often convoluted strategy may have had a purpose at one time, it’s counterproductive now.

We often see complexity as showing sophistication. A simple approach, it’s assumed, can’t be productive. The problem, of course, is that a complicated approach often works against you. It waters down your budget, drives up costs, and hurts your results.

Your goal shouldn’t be to follow everything I’ve said here blindly. It should be to thoroughly understand how everything works so that you’ll know when your involvement is necessary and when it’s a hinderance.

And then execute it consistently.

Your Turn

Anything you’d add to the Common Sense Playbook to Meta Advertising?

Let me know in the comments below!