Facebook Interest Targeting:
The Definitive Guide (2021)
This is a complete guide that covers every aspect of Facebook interest targeting.
- How to find interests.
- How to pick the right interests.
- Facebook interest targeting tools.
- Interest testing strategies.
- Scaling interest-based campaigns.
- And lots more...
If you want to level up your Facebook interest targeting skills, you'll love this guide.
Let's dive right in.
Interest Targeting Basics
Picking The Right Interests
How To Find Interests
Interest Testing Strategies
Interest Scaling Tactics
Facebook Interest Targeting Basics
In this first chapter I'll cover the basics of Facebook interest targeting.
You'll learn what Facebook interest targeting is and how the Facebook advertising battlefield relates to the story of David and Goliath.
I'll also highlight the importance of proper interest research (especially for smaller businesses) to stay ahead of the competition.
Even when Facebook ad costs keep rising.
What is Facebook interest targeting?
Facebook interest targeting is the process of targeting ads to an audience of Facebook users based on their assumed interests, in order to increase the relevancy of the ad to a target audience that is specifically interested in the subject.
Why understanding Facebook interest targeting is important for advertisers?
A deep understanding of Facebook interest targeting is important for Facebook advertisers of any size, but especially for the smaller advertisers.
Because a limited ad budget means you can't afford irrelevance.
Historically the cost of showing ads on the Facebook ad platform has only risen as a result of the fast growing number of advertisers versus the limited ad inventory.
Recent data shows there are now 9+ million active advertisers on the platform.
Which means you're battling for peoples attention with more than 9 million others!
It is expected that it will only get more competitive and more expensive to reach people on Facebook.
So the most important success factor of any Facebook campaign is to know exactly to whom you want to show your ads.
Your ultimate goal should be to avoid wasting your budget on people that will never buy from you, or worse... people who aren't even interested in what you offer.
Obviously it's not possible to completely avoid reaching people who are not interested in your offer. But you should make it your top priority to avoid waste at all levels in your campaign setup process.
That's why spending time and energy on Facebook interest targeting research, before creating your campaigns will really increase your chances to see success.
It will pay off.
A significant part of this guide is focused on the research related to the audience you're targeting. Because if you're targeting is off, it will never work.
For example, if you're showing an ad for a pet product to someone who doesn't have a pet... it can be the best pet product ad in the world, but it won't bring you any sales.
On the other hand, if you have mediocre ad but you're showing it to exactly the right audience you might see some results.
And if you have a great ad and you're targeting is spot-on... you'll win 🏆
Even when competition increases and ad costs rise.
David ⚔️ Goliath
Are you familiar with the story of David and Goliath and the moral of how underdogs can overcome the odds and be successful?
It's a great metaphor for advertising on Facebook.
As a small business advertising on the Facebook platform, you're David.
You don't have enough budget and not enough data to win against Goliath, in the shape of big advertisers. Not when you're playing their game by their rules.
There's a different set of rules that smaller advertisers should follow, to be able to outsmart giants and win.
The most important difference is that big advertisers have the budget to be irrelevant (a.k.a. show their ads to a lot people who are not interested in their offer).
And they can rely on the Facebook algorithm to find the right people for them, because they feed it loads of fresh data.
Smaller advertisers on the other hand usually don't hit the volume of recent conversion data needed for Facebook's algorithm to optimise effectively...
More is better, in this case.
More data is better. But more data is also expensive to acquire.
If you spend let's say $50 per week on Facebook ads to acquire 10-20 new leads, that can be great for your business.
But you'll never hit the required data treshold where Facebook can predictably and consistently find more leads for you.
Without your guidance on how to actually find them.
Goliaths can just burn half of their budget on irrelevant ad impressions and still see a decent return on the other half, simply because all these "wasted" impressions and the actions that follow will give Facebook valuable data points to optimise.
Their big money buys data 💰
If you're David, you need be hyper-specific when defining the audience you want your ad impressions to be served to. You can't afford burning half of your budget, hoping the other will make up for it.
Because you can only serve a limited number of impressions for your budget.
And you can't fully rely on Facebook to find your leads and customers either. Because you can't afford the recent conversion data it needs to do that consistently.
You need to define the targeting parameters and be smart about it.
That's why understanding Facebook interest targeting is so important.
Especially when you're David...
Now that we've laid the groundwork, it's time to take a Facebook interest targeting deep-dive.
Before finding, picking, testing and scaling interests you'll need to have a in-depth understanding of how Facebook interest targeting actually works.
How Facebook determines interests, how language and location influence results and when you should and shouldn't use it.
When to use Facebook interest targeting?
Facebook interest targeting is one of the available targeting options in Ads Manager.
Next to targeting custom audiences, lookalike audiences or targeting based on behaviours, job titles, employers, life events or geolocations.
You should use Facebook interest targeting when you're just starting out and you don't have a lot of (high quality) pixel data yet.
Then it's your best bet even.
Many advertisers get this wrong.
There is a weird but very persistent belief that Facebook lookalike audiences will always outperform Facebook interest targeting because "the Facebook algorithm is smarter than you".
But that is simply not true.
And it's also very logical why that can't be true.
If you feed Facebook's algorithm shit, then that's also what you'll get in return.
It's really that simple.
If you don't get a lot of (recent) conversions and build lookalike audiences from low-value pixel events like website visits, then your lookalikes will be low-value too.
Facebook won't be able to find consistent patterns in your seed audience and the result will be a fairly random lookalike audience of people that match this weak pattern.
You should definitely use lookalike audiences when your campaign starts to mature a bit and you have a decent volume of value-events, like purchases.
Another reason to use Facebook interest targeting is because it's like market research to learn as much as you can about your target audience.
Using lookalikes and hoping Facebook's blackbox will magically find your next customers is a weak excuse to learn who your perfect audience is.
You need to know this.
You should also use Facebook interest targeting when your audience can be clearly defined by a set of common interests they share.
Like a specific hobby, passion or brand interest. But we'll get to that later.
How are Facebook interest determined?
Facebook uses many data points to determine the interests of their users.
It's a combination of the Facebook pages users like, the content they engage with on Facebook and even off the platform (the Facebook tracking pixel is embedded on millions of websites).
This means you can be assigned an interest because you actually liked or followed an eponymous Facebook page.
But you can also get it when you comment on a friends post about a certain topic.
Facebook's definition of "being interested" is very broad and it might be different from how you perceive being interested in something in the real world.
There you'll find a list of interests that Facebook determined based on your activity on Facebook, such as engagement with pages and ads. It's not a complete list, but it gives a surprising insight into what Facebook thinks you're interested in.
We'll get back to this list a little later in this chapter, let's first look at an interesting example that shows why it's important to distinguish Facebook pages (and the people that like and follow these) from Facebook interest-based audiences.
As you can see in the screenshot below there are 400,168 people who like the Facebook page of Mailchimp, a supplier of email marketing software.
But when you enter "MailChimp" into Facebook Audience Insights (one of the research tools we'll discuss more in-depth later in this guide) the number is very different.
Now Facebook shows that the worldwide audience of people interested in Mailchimp contains 2.5 to 3 million people.
Instead of 400,168 people.
How is that possible?
Well, these people could be assigned the interest because they've visited Mailchimp's website that has the Facebook pixel on it.
Or because they liked a friends photo where she celebrated her succesful migration from another email tool to Mailchimp.
So liking a Facebook page and being "interested" in a topic are different things.
Notice how they're even written slightly different: "Mailchimp" vs "MailChimp".
It's important to note here that the bigger the interest audience is, the greater will also be the difference with the number of people that like the eponymous Facebook page.
For example this Facebook page of CrossFit Mayhem has 100,288 likes.
When I type it into Facebook Ads Manager I find that the worldwide audience for this interest is 129,010 people. Still not exactly the same, but definitely a much smaller percentage gap compared to the Mailchimp example I just showed you.
There are even examples, especially when you look at smaller interests, where the Facebook page has more likes than there are people in the interest audience 🤔
Like this "Radiant Body Yoga with Kia Miller" Facebook page that has 44,186 likes.
While the interest has a worldwide audience of only 18,180 people...
I could give you many more examples, but I don't think these would further clarify the point I'm trying to make in this section.
So let's agree it's weird and move on...
The difference between Facebook pages and Facebook interests
As I showed you in the previous section, there is a difference between liking a Facebook page and being "interested" in a topic (in the eyes of Facebook).
This is a huge frustration for Facebook advertisers 🤯
Because many advertisers would like to be able to target the people who like their competitors Facebook pages, obviously.
But in most cases these aren't interests.
Pro Tip: There is a (paid) work-around that doesn't use interests, you can read about it here: "How to target Facebook ads to fans of a competitor page".
Another frustration that results from this difference, is finding out that the relevant Facebook pages you listed during research can't be targeted with Facebook ads.
The majority or Facebook pages, even the ones with really large audiences, will not show up when you type them into Facebook Ads Manager to target the interest.
Like for example DigitalMarketer, with 431,699 people who like their page.
But there's no targetable interest showing up inside Facebook Ads Manager.
While Ryan Deiss, the founder and CEO of the DigitalMarketer company, has a lot less people liking his Facebook page.
But his name does show up as an interest that can be targeted. With a worldwide audience size of 777,408 people 🤔
OK, so the availability of an interest is not related to the number of page likes.
And there's no magic number of page likes, where you become "famous enough" to become worthy of a targetable interest on Facebook.
Sometimes the name of a Facebook page will show up as an interest and sometimes it won't. Sometimes it will show up.. looking slightly different (Mailchimp vs MailChimp).
Interests and Facebook pages are separate things, but they can be associated.
Meaning that a Facebook page can have an interest with the same name.
But even it that's the case, you'll notice that the audience size of the interest is often completely different (and often MUCH larger) than the number of likes.
There's no clear logic behind why some Facebook pages have associated interests and others don't. It seems completely random and I think only Mark knows...😅
One thing that's unmistakably clear from looking at these examples, is that Facebook heavily inflates the number of people that are "interested" in a topic.
Below you'll find a screenshot of some of my own "interests".
✅ I'm into business, email marketing and social media marketing.
✅ I work from home a lot, so that could explain the "home business" interest.
✅ I like explaining stuff to others, although "teacher" is not my profession.
✅ I'm interested in machine learning.
❌ I'm an Apple guy and would probably never switch to Microsoft.
❌ I don't think I'll ever buy a Tesla either.
❌ Construction is not my thing, I have two left hands.
❌ I think fortune-telling is hocus pocus.
Many people have "interests" in the eyes of Facebook that they don't really have.
Which means that advertisers target "interests" that many people don't really have.
Let's imagine that an online store selling tarot cards and other fortune-telling related products would target the interest "Fortune-telling" 🔮
I would then be included in their targeting, as one of 13,170,450 people in the audience.
But I'm 100% NOT interested in their products and a waste of their ad budget.
Makes me wonder how many others in that audience also are...🤔
This is something that you have to keep in the back of your mind when reading the following sections of this guide on Facebook interest targeting.
Translated Facebook interests
Now let's talk about the role of language when using Facebook interest targeting.
Facebook is used around the world and it is available in a lot of different languages.
Users can select their preferred language so their Facebook user interface will be displayed in that language. There are quite a few languages to choose from.
When you use Facebook with a specific language setting the interest suggestions in Facebook Ads Manager will look different, because these are translated too.
Translated interests are the same interests as in other languages. They also have the exact same audience. They're just translated to the users' preferred language for easier interpretation.
Let's look at an example.
When I have set my Facebook language preference to English (US) 🇺🇸 and search for the interest "Dogs" in Facebook Ads Manager, I find a 564,271,400 audience size.
When I change my language preference to Spanish 🇪🇸 and do the same search (dogs is then translated to "Perros") I still find the same 564.271.400 people audience.
And when I change the language preference to my own language (Dutch 🇳🇱) and repeat the same search for "Honden", I find the same audience size again.
It's the same interest with the same audience, just translated for easier interpretation but it doesn't change the available audience to show your ads to.
In the following example I've switched my language preference back to English (US) and narrowed down the Locations to only show my ads in the Netherlands.
I then typed in the English word "Dogs" (so not the Dutch translation "Honden") and as you can see the total audience size of the interest is still the familiar 546,271,500.
But now the potential reach of my campaign in the Netherlands is 3 million people.
I hope this section demonstrated the limited impact of your language settings on Facebook interest targeting.
You can use any language to find interests. And you can use interests in any language to target any country.
For many Facebook advertisers this is a difficult concept to grasp. They think that interests are somehow related to a specific country, which most of them aren't.
Think about it like this: there are many people around the world who are interested in "Dogs". This interest keyword can be translated to many languages depending on your Facebook language preference.
But it's still exactly the same audience of people who are interested in 🐕🐩
Narrowing down your ads reach by geographical location will change the reach of the campaign, but it doesn't change the total audience size of the interest.
There are also "local" Facebook interests.
Local Facebook interests
I just showed you how the same interest in different languages still has exactly the same audience size, these "translated" interests form the majority of all interests.
But there are also "local" interests, meaning that the keyword will not show up when searched for with a different "locale" setting.
I'll get back to this and show you how you can change this "locale" in the next section of this Facebook interest targeting guide where I'll discuss how to find interests.
For now it's enough to know that if you use Facebook from the Netherlands like me, the locale will be set to "nl" by default. And that's why I'll see Dutch local interests suggested when I search for a Dutch interest keyword.
Local interests can be things like local stores or brands.
I'll show you an example.
I still have my language preference set to English (US) and I've selected "Worldwide" as the location to show my ads. I'm working on my laptop in Gouda (the Netherlands) so by default my "locale" will be set to "nl".
When I enter the Dutch keyword "vakantie" (vacation / holiday) I get both English as well as Dutch keywords as suggestions.
It will always be a mix of both when you search for a keyword in a different language than your Facebook language preference.
Keywords like "Vacations", "Cottage" or "Holiday cottage" are translated interests. When I switch my language preference they will be translated to Dutch.
Vacations is then "Vakanties" with the same 330 million people audience.
Keywords like "VakantieVeilingen", "Roompot Vakanties" and "Veluwe Vakantieparken" are local interests. These are local brands, just known in the Netherlands.
As said earlier, most interests on Facebook are just translated to the users' preferred language.
But there are also local interests that will only show up depending on the "locale" so they're scarce and also a bit harder to find.
I'll cover finding interests in the 4th chapter of this Facebook interest targeting guide.
Picking The Right Interests
Facebook interests are NOT born equal...
In the previous chapter of this definitive guide on Facebook interest targeting, you learned that Facebook heavily inflates the number of people that have an interest.
The result is that many people in these audiences are not really interested...
In this chapter you'll learn what makes a good interest and how to pick the right ones.
The "but nobody else criterion" for picking the best Facebook interests
If you're selecting interests for your Facebook advertising campaigns, you should aim for interests that clearly define and isolate your target audience.
You can do that using the "but nobody else" criterion
I'll use an example to explain this criterion.
Let's say you have a Golf store and sell different types of products to enthusiastic amateur golf players. So the audience you would like to reach with your Facebook ads are people who regularly play golf.
The worst thing you can do is exactly what Facebook recommends you to do... target the broad interest "Golf".
While it sounds silly, that is actually NOT an interest you should go after.
At least not from the start.
When your Facebook campaign matures and you've collected a lot of conversion data you can gradually go after broader interests because Facebook will then be able to find segments within the huge "Golf" audience that are most likely to convert.
We'll discuss scaling interest-based targeting in the final chapter of this guide.
You definitely shouldn't start your campaign targeting broad interests like "Golf".
The first problem is that most advertisers target these broad audiences because they don’t know better and Facebook suggests it to them. The second problem is, they have huge audiences with many people who aren't even mildly interested in the subject.
Think about it... how clearly defining is the interest "Golf" for true golf enthusiasts?
Not at all.
There are MANY people Facebook thinks are interested in Golf. Based on their page likes, website visits, comments on friends posts about golf, etc. But that doesn't mean they're actually interested in Golf and likely to buy your products.
Facebook likes to inflate the size of these audiences as much as possible, because that increases the number of impressions they can sell and their revenue.
But these broad interest audiences include a large percentage of people that ended up in there by accident. You don't want to show them your ads.
You want to show your ads to people who REALLY like golf.
You might want to target 'Tiger Woods', the most famous Golf player of our time.
That might be a better idea right?
There are MANY people that are "interested" in Tiger Woods in the eyes of Facebook, that might not be golf players at all. Tiger has been all over the media for a lot of things that had nothing to do with golf (like cheating on his wife for example).
He also was the face of consulting firm Accenture.
Are all these people that interacted with Tiger Woods content enthusiastic golf players and the perfect audience for your golf store? Probably not.
Picking the right Facebook interests means you have to think about the interest and assess if having that interest is a defining characteristic of your perfect audience.
Or do other people have this interest too that are NOT your audience?
In other words, does it clearly define and isolate your perfect audience?
A great interest for your golf store is the interest "Bubba Watson".
If you're not into golf you have probably never even heard of his name. But if you are enthusiastic about golf, you'll know his name for sure.
His name is actually Gerry Lester "Bubba" Watson Jr., so it's the nickname of a multiple Golf Masters Tournament champion.
This is a perfect interest to target, because your target audience of enthusiastic golf players might have this interest "but nobody else" will.
That’s the big idea behind the “but nobody else” criterion.
It's defining your perfect audience and isolating everyone else.
Keep this criterion in the back of your head when doing Facebook interest targeting research.
It’s not something that you can 100% apply all the time. But the idea of a hyper-clear audience definition and isolation of everyone else, is the key to targeting the right people while limiting waste of your ad budget.
How To Find Interests
Now you know what to look for when selecting the best Facebook interests to target, it's time to start actually finding these interests.
In this chapter I'll cover various Facebook interest targeting tools and methods to help you find a long list of interests to choose from.
You'll even learn a trick that most advertisers don't even know exists, using the so-called Facebook Marketing API (without tech skills).
This will reveal interest others can't find...
We're all used to Google to search for and (hopefully) find things.
But when it comes to finding Facebook interests to target, I would recommend not to use Google too much. It will save you a lot of disappointment and frustration.
That's because Facebook and Google are different platforms.
Google's "keywords" and Facebook's "interests" are a different beast.
When you do use Google to find keywords related to your product or niche, you'll often find that these keywords don't exist as Facebook interests that can target with ads.
So you'll be tossing in keywords like a slot machine praying it will return some other response than "no results found" 🙏
On Google it's all about semantics and finding plurals, synonyms and combinations of multiple keywords (long-tail).
On Facebook there's a list of specific interests that can be targeted, and you'll have to know what they are or search for them.
I'd highly recommend to just use search tools that are using Facebook interest data, so there's a high probability of finding interest keywords you can target with ads.
Facebook interests list
There are many Facebook interests you can target.
But the list constantly changes, with interests being added to it and removed from it.
Sadly Facebook's doesn't provide a full (master) list of all the different interests that you can target at any given moment.
They only share a static list of broader Facebook interest categories.
You can see this list when you go into Facebook Ads Manager and then click on the Detailed Targeting field you'll find the option to browse through "Interests".
You'll quickly notice it's a fairly short list and most interests that are available for targeting will fall outside of these categories.
So please don't put too much weight on the Facebook interests list.
It can be used for a little bit of inspiration and to spark new ideas, that's all.
Facebook Ads Manager
We'll stay inside the Facebook Ads Manager, to discuss how most advertisers search for interests to use in their campaigns (when they use Facebook interest targeting).
It's the interest suggestions list that shows up when you start typing interests.
It will show a list of around 25 ideas for other interests you can target, related to your input keyword. This list also includes job titles and behaviours, so it's actually a little less than 25 interests.
Many advertisers use this thing like they also use Google.
They type "fitness a", "fitness b", "fitness c" in hopes of getting different suggestions each time. In most cases it will be duplicates with just a few new interests.
Searching interests through this tool can be time-consuming, because you're basically guessing input keywords hoping to find relevant related interests.
One thing you might notice is that the interests that Facebook suggests in this dropdown are mostly interests with a large audience size.
Facebook has determined there are a lot of people interested in these topics.
As you know from the previous section, that isn't always a good thing...
Facebook Marketing API
Most web applications with a broad audience try to constantly simplify their UI.
Facebook's ultimate goal (for the Facebook Ads Manager at least) is to make it as easy as possible to create an ad. It needs to be dummy-proof, so everyone can do it.
As a result they've removed many buttons, options and more advanced features and data from the interface. They also closed down "Power Editor" which was meant to be the interface for more advanced users.
But that doesn't mean the options and data are gone.
Facebook offers clearly documented API's that developers can use to build helpful software and tools around Facebook's data and functionality.
One of these API's is the so-called "Marketing API".
This is what it looks like:
As you can see it shows interests as well as the (worldwide) audience sizes.
It is the source of data that feeds the interest suggestions dropdown that you saw in the previous section. But it doesn't limit the displayed results to just 25 suggestions.
It just shows all the available raw data.
It searches for interests related to the input keyword "Fitness" and returns hundreds of related interests in a second, while Facebook Ads Manager shows just 25 ideas.
The only way to find these interests otherwise would be to guess them by entering dozens of input keywords hoping to discover related interests in the suggestions.
Pro Tip: Change the keyword in between brackets in the API URL to run searches for different input keywords. It works best if you use broad input keywords. If you'd like to learn more about this Facebook Marketing API, I recommend to read my popular Medium article on this topic.
This trick reveals a lot of hidden interest data, but it's not very usable. And to many Facebook advertisers the JSON code doesn't look very attractive either.
That's why I've created InterestExplorer, a search tool that uses Facebook's API to reveal hidden interests others can't find inside Facebook Ads Manager.
This Facebook interest targeting guide is clearly not just about the software I've created, but it does deserve a section in this chapter on finding Facebook interests.
When using InterestExplorer you basically automate the process of using the API and processing the results. You can just enter any broad interest and then hit Explore 🔍
Because it uses the Facebook API in realtime you'll only find interests that can be targeted with ads right now. I'll quickly list some of the other features here:
- Click on the interests you like to add them to your selection box.
- Copy interests to your clipboard, save them to a project or export to a CSV.
- Change the "locale" of your search to find some scarce local keywords.
- Use the Google and Facebook icons to run quick-searches for the interest.
You can just copy the interest you've found into Ads Manager, like you would normally also input interests. They will all be accepted for targeting, just not auto-populated.
So you'd have a hard time finding them...
It's a simple tool, but it's a huge time-saver during Facebook interest targeting research. It will quickly show you interests that are hidden to most advertisers.
So you can target these and avoid most of the competition.
These smaller interests are often laser-targeted, but also a bit too small to create a campaign around. That's why later in this Facebook interest targeting guide, you'll also find a chapter about creating audiences the right way (balancing relevance and size).
Nobody has just a single interest.
Your perfect customer will have more than one interest related to the niche you're targeting.
And the most valuable segment of your target audience will be the part where multiple highly relevant interests overlap.
In this chapter I'll cover the best ways to create an interest persona and create your audiences.
And how to avoid the most common pitfalls!
Creating Facebook interest targeting audiences in Facebook Ads Manager
There are multiple ways to create audiences for Facebook interest targeting.
My preferred method is to keep audience creation separated from creating the ad campaigns. This keeps things organised and allows to focus just on audience creation before actually setting up the budget, placements and ad creative.
Inside Facebook Ads Manager, you'll need to open the menu and then Audiences.
In there, you'll need to click on Create Audience and then Saved Audience.
Then in the modal screen that opens, you can set up all parameters of your Facebook interest targeting audience. Like the location, age, gender and language of the people you'd like to target with your ads. If they match your selected interest(s).
You can also create the audience when you're creating the ad sets and ads.
But I would highly recommend to separate audience creation from the other steps.
This allows you to focus just on the audience, map out multiple audiences and have the ability to go back in and edit them at a later moment - in one centralised place.
Alternatively you can also create Saved Audiences from within Facebook Audience Insights. You can also open your Saved Audiences there to analyse the audience.
Single interests vs interest profiles
Before you build out your Facebook interest targeting audiences, it’s important to realise there’s a big difference between an interest profile vs. single interests.
Nobody is defined by just a single interest.
Every person has many different interests and the combination of these defines their overall interest in a topic. So it’s best to think about a profile that contains many interests that define and isolate your ideal customer.
You can visualise the fact that people have multiple interests, by creating two or more (single interest) Saved Audiences and then clicking on Show Audience Overlap.
Overlapping circles will then visualise the number of people that have both of the interests you've selected.
In the example below you'll see that 310,000 out of the 16 million people that are "interested" in Tiger Woods are also interested in Bubba.
That's only 2%.
If you own the Golf store from the example, would you then rather target these 310,000 people that are interested in both golf players?
Or just the huge audience of 16 million people that are "interested" in Tiger Woods?
Where do you think your (limited) budget would be best spent?
Many of the Facebook ads gurus out there will tell you that you should create many ad sets with different interests when starting out. To test which interests work for you.
Of course this might work in some cases, but there is a much better way to start out…
The problem with targeting single interests, meaning 1 interest per ad set, is that the success of the ad set will largely depend on the size of the audience.
If you target a highly specific single interest, the ad set might perform really well.
But if you target a broader interest by itself in 1 ad set, it’s a huge audience and Facebook will have to guess who the best users are within that audience.
Most small advertisers can’t afford the wasted impressions that Facebook shows, because it lacks the data to find the right people within the larger pool.
And even if you do get a conversion here and there, it’s best to attribute that to just pure coincidence...instead of drawing the conclusion that it is a winning interest...
Creating an interests persona
As mentioned before, nobody has a single interest.
Your perfect customer will have more than 1 interest related to the topic/niche you're targeting. The most valuable segments of the audience will be the areas where the circles overlap.
Especially when these are "but nobody else" interests to start with.
I like to call this an interests persona.
As an advertiser it’s crucial to build out an interests persona, a profile of your ideal customer as defined by their interests.
And then work on that profile and improve on it, based on research and learnings.
It can be helpful to categorise the interests you’ve found during your Facebook interest targeting research, into various categories.
For example: brands, magazines, public figures, websites, etc.
I usually create a Google Sheet for this and add these categories as column titles. Then I’ll paste the interests I’ve found in the columns underneath the different categories.
So all the brands in the brands column, all the public figures in the public figures column and so on. This helps you get an overview of the different types of interests you have collected during your research.
By doing this you'll get a clear picture of the ideal person you’re trying to reach.
After categorising the interests, you can dive straight into the audience creation process to start testing these interests and their combined interest profile.
But personally I like to make it more visual for myself, so I create a persona slide for the ideal person I’d like to reach with my campaign. This helps ME, but it can also be very helpful for your team members or in the communication with your clients.
In the center of the slide I place a picture of my ideal customer.
I give him or her a name and an age as well.
My goal with this persona is not to be 100% complete or comprehensive. Picking a guy of a certain age, doesn’t mean I won’t target females or other age ranges.
It’s just defining who’s the person MOST likely to purchase our products.
That’s whom I want to start with.
To see if my ads convert at all.
Later on I can use the data Facebook gives me to segment further.
Around the picture of my ideal customer, I add buckets of interests in the different categories. I usually picture these as speech balloons, saying for example: “I like to read these books/magazines/websites”, “I like these brands”, etc.
Print it out, stick it to your wall, this is WHO you're are going after.
Pro Tip: A final (optional) step is to validate and enrich your interest persona with existing customer data you might have. Maybe you have already sold a lot of products. In that case you can check your order system and sort it by the highest value customers you have, search for these people’s Facebook profile and study it to learn more about your ideal customer. Which pages do they follow, what posts do they comment on, what images do they share, what kind of language do they use? Even if you can’t use this input as new interests to target, you’ll get a lot of inspiration for you marketing hooks, ad copy and creative.
Interest stacking and layering
When you acknowledge the fact that your ideal customer has more than one interest, there are multiple ways to approach Facebook interest targeting.
- Targeting a single interest in one ad set (not great to start with, unless it's a highly specific interest that clearly defines and isolates your ideal target).
- Adding multiple interests in one ad set to target their summed total audience size, this is what's called "interest stacking".
- Adding multiple interests on one ad set to target narrow down your audience by just targeting the overlap between interests. This is called "interest layering".
I'll discuss both the interest stacking and layering strategies in more detail.
"Stacking" interests means you add multiple interests into the Detailed Targeting field in Facebook Ads Manager. The resulting audience is the sum of the separate audiences of these interests.
So it adds up.
In the screenshot below I've stacked 6 interests into one ad set's targeting field to target the combined audience of all these interests and their overlap.
The result is a huge audience of 350,000,000 people.
Let's quickly look back at the screenshot I shared earlier with the single interest Saved Audiences for each of these interests.
As you can see just the "Golf" audience on its own is also 350 million people.
Which means all the other smaller circles fall within this huge circle.
Adding Titleist, Bubba Watson, Tiger Woods, Golf Channel and Golf Magazine to the stacked interest audience, doesn't add a single person to my target audience.
Because they 100% overlap with the broad "Golf" audience.
So "stacking" interests isn't meant to be used like this.
It's meant to target specific interests that would be too small to go after as single interests. That's when you should stack them together into one bigger audience.
Pro Tip: Going after these all-encompassing interests (like "Golf") is a great way to scale your campaign when it has matured a bit, and it has given Facebook a large volume of conversions to exit the learning phase and find more of these people for you within the extremely large audience.
The other option when targeting multiple interests in one ad-set is called "layering".
This is a little-used method to better reflect your interest profile and to basically force Facebook to show your ads just to the best segments within your target audience.
Layering is my personal favorite.
The idea is simple. Would you rather show your ads to people who are interested in ‘Bubba Watson’? Or to people who are interested in ‘Bubba Watson’ AND in the golf brand ‘Titleist’ AND also "Golf Magazine"?
Nobody has just a single interest. Everyone has many interests, and the combination of interests (the interest profile) defines AND isolates our perfect customer.
That’s where layering comes in.
In Facebook Ads Manager, when you input at least one interest in the Detailed Targeting field a small option appears underneath that field. The link says ‘Narrow audience’ and when you click it a second Detailed Targeting box appears.
On top of that box it says ‘and must also match’.
Now our layered targeting is setup to target just the audience that is interested in Golf Magazine AND also is interested in Titleist.
So it's the overlap between these.
I can repeat the same step again by clicking Narrow Further to add another layer to my targeting.
There I'll enter Bubba Watson to target just people who are interested in Golf Magazine AND in Titleist AND in Bubba Watson.
Pro Tip: The order in which you input these interests into layers doesn't matter. It will be the same audience when you would put the smallest interest at the top.
Obviously the remaining audience will be smaller than targeting just the single interests, but also laser-targeted. And in this case still a 230,000 people audience.
I'd much rather start my campaign targeting these Golf enthusiasts that have multiple highly relevant interests, instead of going after huge interest audiences.
Think about it; even when you would spend just $10 CPM (the cost to show your ad a 1000 times) it would already cost you $2300 to reach all these people just once!
I know where I would put my money 💸
By layering you’ll completely exclude people who accidentally ended up in an audience. You’ll only target people who match multiple highly relevant interests.
There is one challenge when it comes to layering interests; audience size.
The audience size can get too small when you'll layering interests and you're also narrowing down the reach by targeting a smaller geographical area.
Like in the example below where I just target the Netherlands which leaves a potential reach of fewer than 1000 people...
This is just the reality of advertising on Facebook.
When you target a huge geographic area you can be very specific with interests. When you target a small geographic area you should be less specific with your interests.
You can't have both.
Here's what you do:
- You can reduce the number of layers, so instead of 3 you'll just use 2.
- You can add in a few broader interests to increase the audience size.
It does become even more important to target just highly relevant interests that clearly define and isolate your ideal customer.
You'll always need to balance audience size and relevancy.
So when is an audience too small?
More than anything else, that's a personal decision.
Personally I don't have a problem with creating smaller ad sets if they are highly relevant and perform. But you might get in trouble because the campaign will quickly run out of fresh blood.
More importantly, you can't spend the same budget on ad sets with a 10.000 people audience vs an audience of 10 million people.
A good metric to look at is the frequency.
This is the average number of impressions a user in the target audience has seen. It is an average, so there will be people who have seen it a hell of a lot more…
Personally I aim for audience sizes of at least 50.000 people. There’s no real reasoning behind that, but in my experience anything smaller than that really quickly fades out.
The maximum budget I spend on an ad set per day, is $10 per 10.000 people.
That’s a very simple rule to make sure you distribute budget across ad sets in a balanced way. With CBO campaigns this is different, but there I still look at the audience per ad set and then sum the max. budget per ad set to find the maximum budget for the campaign.
Interest layering and stacking combined
When you've found many highly relevant but smaller interests during your Facebook interest targeting research, you can also create an ad set that combines both the stacking and layering strategy.
You'll then create 2 or 3 layers and add multiple interests into each layer.
Within the layer the OR relationship will apply, meaning that someone in your audience should have at least 1 of the stacked interests in that layer.
Between the layers the AND relationship will apply, meaning that someone in your audience should have at least 1 interest from each layer.
You can just randomly put these interests into the layers you've created.
Our InterestExplorer software also has a built-in layering feature to help you do that.
When you've created a project with relevant but smaller interests, you can click on the "Create layers" button and create the number of interest layers you'd like to use.
Then you can randomly assign the selected interests to the layers.
You still have the option to manually drag-and-drop interests to different layers if you like. And then copy each layer to your clipboard to paste them straight into the input fields in Facebook Ads Manager.
In a few seconds it's done 👇
Now you know how to create Facebook interest targeting audiences the right way, I'd like to end this chapter with two important warnings.
Facebook detailed targeting expansion
When setting up your targeting during the creation of an ad set, Facebook shows a little checkbox to activate Detailed Targeting Expansion.
If you activate it Facebook will "show your ads to more people which may help your reach your optimization goal".
TLDR: they'll show your ads to more people.
In the example above it will increase the 500,000 people audience size for the interest "Bubba Watson" to a 2,100,000,000 people audience.
Don't use it, it’s a trap ⚠️
Over time I have run many experiments to test this Detailed Targeting Expansion feature. In theory it should be the ideal method for scaling your relatively small interest-based audience to a bigger reach.
I mean… what if Facebook could find more relevant interests to target, based on the initial set of interests you’ve chosen.
Sadly, I’ve never seen it improve the performance of any of my campaigns.
A second issue I have with this feature, is that it creates a total black box as to whom you’re showing your ads to.
I’ve put a lot of work in truly understanding who my perfect customer is and what his or her interests are. I only want to throw that overboard, when Facebook would really be able to increase my performance. That’s has never been the case.
It’s just a way to expand the reach of your campaigns and their revenue.
Then there's another important warning.
Always make sure to exclude your existing website visitors from acquisition campaigns.
This is a mistake I see way too many inexperienced advertisers make… NOT excluding their past website visitors. Most people exclude only their past purchasers.
I highly recommend you to separate your acquisition campaigns from your re-engagement or retargeting campaigns.
The reason for that is that if you don’t do this, every campaign is a retargeting campaign at the start.
Your goal with an interest-based campaign is to reach cold traffic that matches the profile of your perfect customer, based on their interest profile.
When you start that campaign thinking it’s an acquisition campaign, Facebook will use the data it has to try and find the best people within your target audience to show your ads to. One of the most important data points Facebook has is past website visits.
People that have already visited your website before and match your interest profile are the single best segment to show your ads to.
So that is what Facebook will do 💡
But that is not cold-traffic acquisition.
This is the reason why campaigns might start off great, getting conversions on the first day. It’s because Facebook retargets your past website visitors first. While you think you’re targeting a new audience. And then when the retargeting pool is depleted, it will go after the cold audience and naturally the campaign performance will decline…
Don’t make this mistake and always exclude website visitors in the past 180 days from your acquisition campaigns. That’s the longest timeframe you can choose.
Interest Testing Strategies
When your research is completed and you've found a long list of relevant interests, it's time to put these to these test.
In this chapter of the Facebook interest targeting guide, I'll discuss various ways to test using both CBO, ad-set budget and AB-tests.
But I'll also explain why the majority of advertisers doesn't really understand testing,
And the point of doing any tests at all...
Don't test at all...
When you’ve collected a list of interests that clearly define and isolate your ideal target, it’s time to put these to the test. There are various approaches you can choose from.
If you want to test them at all...
That’s the first point I’d like to discuss.
As a Facebook advertiser, you’re goal is to make a profit. To make a positive return on your ad spend. You’re not a scientist who's life goal is statistical signficance 🧪
So while testing can be very insightful, my advice is to attain a highly pragmatic attitude towards testing interests against each other.
That having said, for years there hasn’t even been a way to test Facebook interests with the goal of drawing statistical significant conclusions.
Why Facebook interest testing is pointless (mostly)
Firstly, to say anything meaningful about the results of an experiment you need to serve the experiment to a large number of people (called "sample size").
If you run experiments with just a small budget, you won't get enough conversions to test the performance of different interest-based audiences towards that goal.
Which means you can just test without significance and the results can be just pure coincedence. For example, nothing in the screenshot below declares a "winner".
Pro Tip: You can use an A/B-test calculator like this one to enter the numbers and see if the difference in performance is statistically significant or just coincedence. I've entered the impressions and purchases from the screenshot above. The AB calculator then shows the result is not significant and I need to collect more data.
Secondly, to say anything meaningful about the results of an experiment you need to make sure that the audiences in your experimental setup don't overlap.
If a person is part of more than 1 group that messes up the ability to draw any conclusions.
Facebook interest audiences overlap, as you've seen earlier in this Facebook target guide. Which means that users are often in more than 1 audience.
That makes it practically useless if you're looking for a "winning" interest and discard the other interests as losers.
And thirdly, to say anything meaningful about the results of an experiment you need to make sure that the different variations in your experiment (interest in this case) get an even share of the spend and preferably an even number of impressions.
All testing methods in Facebook except for the true split-test don't spend your budget evenly across the different audiences. Which means results will also be skewed.
My goal is not to discourage you to test.
Personally I test a lot, but mainly other things than targeting. Like different bidding mechanisms, placements, my funnel and my offer.
I trust in the research that I've done and the audience I've defined using the targeted interests that clearly define and isolate my target audience.
Then I apply layering and stacking to create a target audience that balances relevance and size.
And then my next goal is to find an ad that works for this audience.
Think about it like this; if you've clearly defined your targeting you can be sure that you're showing your ads to the right people. If your ads don't perform when shown to this perfect audience, you'll need new ads.
"Testing" interests with CBO
I've put "testing" in quotation marks, because it's not a clean test.
Before, advertisers mistakingly thought that creating many ad sets with just 1 interest per ad set (single interest ad sets) was an experiment.
Many still think like this by the way.
The ad set with the best stats would show which interest worked best right?
No, false ❌
Ad sets within the same campaign would be shown to the same audience, so there would be overlap. A user might have seen the ad based on multiple of these interests, while his purchase was just assigned to one of the ad sets.
That's not a clean test to find a "winner" and discard the others.
Then Facebook introduced campaign budget optimisation (CBO), where they automatically adjust the budget each ad set gets based on its performance and the total campaign level budget - with the goal of maximising conversions.
The algorithm picks a "winner" really fast and spends most budget there.
Which, as you now know, is not a clean test.
But your goal, and Facebook’s goal with campaign budget optimisation align closely.
You both want to find which targeting works best and assign most budget there. So with a pragmatic attitude you can use a CBO campaign as a testing mechanism.
I usually create a separate CBO campaign as a testing ground (if I test at all).
You can do the same, by creating a separate campaign and then creating single-interest ad sets in that campaign. Let it run for a few days and see which interests get the most budget / reach etc.
Note that you do need to give Facebook time to optimise.
At the start it doesn’t have data to optimise for, so it will just assign more budget to interests with larger audiences.
When it does get enough data to exit the learning phase (50 conversions per week per ad set at least), it start to distribute the budget towards the ad sets that perform best in the campaign.
It’s not a clean test, but it’s one way you can learn about interests performance.
Most smaller advertisers will never get to 50 purchases per week per ad set.
Which means "testing" with CBO completely misses the point.
Don't get me wrong, I run almost exclusively CBO campaigns with interests of which I know that they work well for me. And I also run CBO campaigns with ad sets that target various lookalike audiences.
I just let Facebook decide on the distribution of budget across the different interests, but I don't have the intention to find a "winning interest".
To me all interests that clearly define and isolate my ideal customer are winners 🏆
"Testing" interests with AB
AB stands for ad-set budget.
When Facebook launched Campaign Budget Optimization they communicated that they would sunset the option to set budgets at the ad-set level.
After postponing the last day a few times, they've now decided that both options will remain available. So both setting budgets at the campaign level (CBO) and at the level of individual ad-sets (AB).
Which means you can still structure ad sets the "old-school way" and create for example 10 ad sets with 1 interest in each ad set.
You do have to acknowledge the fact that these interest audiences overlap with each other, so it's not a clean test.
But at least Facebook will try to spend the individual budget of every ad-set. While with CBO they'll pick a favorite early on and not all interests will even start running
⚠️ Don't fall into the trap of thinking that single interests with a few conversions are better interests than others that don't have conversions yet.
For most small advertisers the numbers are way too small to draw any conclusions like that. There will be overlap and too much inconsistency across different days.
Run a Facebook A/B test
The only clean test on Facebook is an A/B test.
When creating a campaign, you can check the A/B test checkbox and then select that you want to create a test around the variable “Audience”.
Then in the process you can select multiple single interests as the audience variations.
Over time Facebook will then be able to show you which audience, so which interest performed the best.
But… that needs time, and budget.
Because every interest audience will get an even percentage of the budget and it needs a good volume of impressions and conversion data to draw a conclusion.
In the screenshot below I've setup an AB test to test the interest "Bubba Watson" against "Phil Mickelson". To run this test with chances of seeing a statistically significant outcome, Facebook recommends to run it for 4 days at $1000 per day.
The budget will be spread evenly at $500 per day per interest.
So it either needs a high daily budget or needs to run for a longer time, depending on the number of audience variations you add.
If you lower the budget, you'll need to run the experiment longer.
If you lower the number of days, you'll need a higher budget.
In both cases you'll end up spending quite a bit to run a decent AB test on Facebook.
I do use split-tests sometimes, but usually to test placements or bidding mechanisms in a clean setting.
For audiences, I chose to be more pragmatic and trust my research, the layering and the stacking strategy to make sure I reach the right people 🎯
Interest Scaling Tactics
You've found interests, picked the right ones and then got it to work (with or without actual testing of the individual interests).
That's the point at which you're ready to scale the reach of your campaign.
Most advertisers get this the wrong way.
They start with a focus on "scaling" and think there is a magic recipe for it...
This chapter will give you a down-to-earth approach to scaling interest-based campaigns.
Start small to grow big
Scaling is the most over-hyped topic in the Facebook advertising realm...🚀
If you scroll through any of the Facebook groups on the topic of Facebook advertising, you'll quickly notice that everyone is looking for the secret recipe to scale.
Often before they even found anything that works for more than a few days.
You can't run before you can walk.
I'm a strong believer in starting small to grow big.
When it comes to Facebook advertising there are a couple of important benefits to starting small. While it might not be "sexy" it's definitely effective, because...
Firstly, starting with a small but specifically defined audience will make sure the right people see your ads and hit your website.
This will make sure that your Facebook pixel events (like ViewContent, AddToCart and Purchase) are triggered by a high-quality audience. So Facebook's algorithm will better understand who you're looking for to eventually find similar people inside your bigger targeted audience.
Secondly, starting with a small but specifically defined audience will make sure you learn faster.
If you're sure your ads are shown to your ideal audience and still they don't perform, you'll know that you need to tweak your ads or your offer (or both).
You can't draw this conclusion if you're not sure if you ad is even shown to the right people.
Many advertisers worry too much about the scale of a smaller target audience. They prefer to go right after the big audiences ignoring the facts I've stated above.
And then they're basically sailing in the dark.
How do you know if it's your targeting, your ads or your offer if you're not even sure if your ads are shown to the intended target audience?
The point I'm trying to make is that smaller advertisers need to embrace the fact that they need to start small to build towards a bigger campaign gradually.
"Scaling" is just adding more budget (often referred to as vertical scaling) or creating more unique reach (known as horizontal scaling).
I'll explain my strategy for both 👇
When I've created a layered or stacked ad set that performs at a profitable ROAS average across 3 or even 7 days I start to think about increasing the scale.
My first step is always to try and see if I can increase the unique reach of the campaign (i.e. to show my ad to more people), by duplicating the campaign and segmenting.
⚠️ Rule #1: Changing Facebook ads almost always breaks them.
So don't do that.
Create new campaigns, new ad sets or duplicate existing ones. And then leave these ones untouched as well. Don't constantly make changes - it's a recipe for failure.
There are multiple ways to scale horizontally.
Decrease the number of Facebook interest targeting layers you use.
This will increase the audience size that your ads will shown to. You can gradually change the balance in the favour of audience size over hyper-relevance.
But with 2 layers and the right ("but nobody else") interests you'll still have the benefit of excluding most of the people that aren't really interested in your topic.
In the screenshot below you can see that my 3-layer interest audience has a potential reach of 3.3 million people.
Just using highly relevant golf interests.
I can scale the reach of this ad set by decreasing the number of layers from 3 to 2. That will increase the potential reach to 12 million people.
Create new ad sets with broader interests.
As your campaign matures, Facebook gradually collects more data about the users that opt-in, add products to their cart and hopefully convert.
Facebook's algorithm will use this data to find similar people within your interest audience to show your ads to. It will try to find the people within that audience that are most likely to convert so they can show your ads to these people first.
There's no specific number of conversions to aim for, but in my experience starting from a few hundred purchasers it makes sense to start testing both lookalike audiences and broad-interest based audiences.
You can either add a really big interest (Golf) and still add a 2nd layer with highly relevant "but nobody else" interests to balance reach and relevance ⚖️
Or even test going a step further and target the massive "Golf" audience.
Create new ad sets to segment demographics
You can break down a single audience into multiple ad sets with their own budget, to force Facebook to spend all these individual ad-set level budgets = reach more people.
- Create audience segments for male and female (targeting the same interests).
- Create audience segments for age ranges.
- Create audience segments for various countries in your targeted geolocation.
Just by using these horizontal scaling tactics you can easily 10x the daily ad budget you spend. And as a result have many more than 1 ad set that performs profitably.
If you've exhausted your creativity in this area, you can still always increase budget.
The same rule applies as with horizontal scaling; don't touch running ad sets.
Facebook recommends to keep your budget increases or decreases to 20% or lower per day to avoid resetting the learning phase of the campaign.
In my experience you can definitely make bigger changes when you increase the budget at the campaign level (CBO) compared to the ad-set level (AB).
But still I rarely do it.
Simply because in most cases making changes to running ad sets will break their momentum.
If you have an ad set or campaign that works well, just duplicate it so it starts as a new campaign. You can then set the budget at whatever higher amount you like.
It's much better (and more stable) to have for example 5 ad sets that spend $50 each, instead of 1 ad-set with a $250 budget.
Don't put all your eggs in one basket 🧺
You've reached the end of this definitive guide on Facebook interest targeting.
I've covered everything I know about finding interests, picking the right ones, testing and scaling your interest-based campaigns.
I hope you enjoyed this 11.000+ word guide and you've learned new things you didn't know yet.
If this content was valuable to you I would like to ask you to help by sharing it on social media.
That would be really appreciated 🙏