3 Times You Should Prioritize Remarketing Over Meta’s Algorithmic Ad Targeting

There was a time when most of my ad budget was spent on remarketing in one form or another: Website visitors, email lists, followers, post engagement, and more. I discarded most of these in favor of Meta’s algorithmic targeting, but there are exceptions.

There are times when remarketing continues to make good, smart sense.

You misunderstand my intention. I still think marketers use remarketing too often. Not only does it not work the way it used to (and marketers often misinterpret the effectiveness of their remarketing results), but it’s also often unnecessary.

Let me explain…

Why Remarketing Is So Important

Don’t confuse the message here. Reaching the people closest to your business is always important.

One of the main reasons that a separate ad set for remarketing is not so necessary now is that algorithmic targeting will prioritize these people anyway. When you use Advantage+ Audience, Meta prioritizes conversion history, pixel data, and previous interactions with your ads.

You can prove this with the help of audience segments. I have seen many times that Meta spends in the range of 25 to 35 percent of my budget on my existing customers and my engaged audience (those who are on my email list or have visited my website, but have not yet purchased from me).

Here’s an example, using Advantage+ Audience without suggestions…

I have noticed this and when I use the first audience it goes more…

Here’s an example using two different ad sets: One using Advantage+ Audience without suggestions and the other using remarketing only.

When using Advantage+ Audience without suggestions, Meta spent 45 percent of my budget on the same people I specifically targeted in a different ad set. By giving the algorithm more freedom, I found that it maintained a reasonable frequency compared to when I only targeted the remarketing group.

Meta now combines remarketing and search to create the right balance. Otherwise it will be more expensive to reach your remarketing audience (which is often and will be the most likely to take the action you want), but the test group is large and cheap.

For this reason, traditional remarketing (where you target broad groups of website visitors, email lists, and people who have engaged with your page) is rarely necessary now. It happens automatically.

Misinterpretation of Results

I should also point out that another reason why some marketers continue to swear by remarketing is because they misinterpret or misunderstand their results. Whenever I see someone sharing conversion results or ROAS that seem too good to be true, it’s usually because the results are inflated.

To be clear, remarketing results it should be good. But they will also be raised. This is a great opportunity to share your results with your test how cute they actually are.

Use the Compare Attribution Settings feature and sort your results by attribution setting. It would also be good to use the First Conversion report (or at least both the First Conversion and All Conversions).

If you remarket, you can expect a disproportionately high concentration in the 1 Day View column. That is usually due to two different conditions:

1. Email people on the same day they are shown your ad.
2. Regular website visitors happened to visit on the same day your ad was shown.

It doesn’t necessarily mean that the ad didn’t do anything. In some cases, these customers saw it and it contributed to their purchase decision. But the most common scenario is that they didn’t see your ad. They were going to buy anyway.

Conversion of views is most valuable when it comes to new customers. They saw your ad or were contacted by it, but didn’t click on it. They remember the brand or product and Google it later. Then they bought.

But if you remarket, at least a decent number of view conversions have not changed.

When Remarketing Makes Sense

Although remarketing is often unnecessary, there are exceptions where it always works.

Here are a few to consider…

1. A specific message for a specific group of people.

A very common example is the abandoned cart situation. You want to show a different ad to people who have just added your product to their cart but haven’t purchased yet. You may want to offer a discount to encourage sales.

Of course, it is debatable whether this is necessary. Meta should prioritize people who have added to a cart when deciding who will see your ads. It will be more expensive to segment those people into a different ad set. Worth checking out.

I’m actually using a variant of this right now. I have a special offer, but I only want a certain portion of my email list to see it. Although it is open to the public, what I love about this offer is the high value people have bought from me before.

This time, I targeted the same people I send emails about this offer. I even refer to the email in the ad copy.

In this way, I understand that the advertisement is only part of the sale. Since it is a high dollar commitment, I hope it will help motivate these people to complete the sale.

I know that my ads will only be partially responsible for the conversions reported to the ad manager. But my hope is to increase the total number of subscribers. Since the audience is smaller, the total amount of money spent on ads will be much smaller, too. And since the sticker price is around $1,000, it’s a low-risk approach that makes sense.

2. Low budget and challenge to get results.

You’re trying to sell a high-dollar product, but you’ve been given a $50 or less budget per day. You don’t have the option to build leads and need to go directly to the auction. Remarketing should be an option.

Of course, remarketing will happen naturally if you aim broadly. But perhaps the advertising audience is relatively small. Even so, you may struggle to get meaningful results.

Remarketing doesn’t guarantee results here, but it’s at least a low-cost option.

3. Top of the funnel optimization.

Optimizing link clicks, landing page views, video views, post engagement, or anything else other than conversions can be problematic. It’s even bigger when algorithmic targeting comes into play because Meta will do everything it can to get you the cheapest action you want. This often results in a loss of quality. With remarketing, you can limit your audience to people you’ve already determined are most relevant.

I did this when promoting my blog posts or Reels. I know I will get a lot of low quality clicks or plays if I let the algorithm search for anyone to interact with. But if my goal is to get more people who have already proven to engage with my content, I will separate them from custom audiences.

Beware of Soft Remarketing

While remarketing still has its place, there is one strategy you should avoid and it goes like this…

1. Run an ad that generates link clicks, landing page views, or video views.

2. Create an audience of people who engaged with the original ad.

3. Identify the people who engaged with the original ad.

The reason this is a problem is the issue we’ve already discussed about optimization. If you’re optimizing for link clicks, landing page views, video views, or almost any other action without conversion, you can expect low-quality work. You create a custom audience for low-quality work. Then remarket to a lower quality audience.

If you are going to use remarketing, make sure you are actually targeting a high quality group of people. Investigate how that audience was created in the first place. Systematically generated audiences or those built in preparation for a conversion will usually be your best bet.

Let Algorithmic Indexing Do Most of the Work

Remarketing still has its place, but you have to let algorithmic guidance do the hard work — especially when you’re getting ready to buy. “Algorithmic targeting” does not include targeting or using Advantage+ Audience. It includes any situation where your audience is growing (and that includes a high percentage of our input now).

Broad targeting should take up the bulk of your ad spend. While there is zero remarketing to people who are already close to you, there is a limited upside. And you want to bring in new people who wouldn’t buy from you if it weren’t for your ads.

Remarketing is a good short-term, low-risk game. Broad targeting is a slow, long-term game that will help ensure you have a remarketing audience to reach in the future.

Your turn

Are you still using remarketing strategies? What specific examples of remarketing success or challenges can you share?

Let me know in the comments below!