This article was originally published in French on Hootsuite blog
The internet is full of advice and secret formulas for Facebook ad targeting. An ever-growing touting of strategies guarantee skyrocket conversion and cheaper Facebook ads campaigns.
Yet, a good portion of this hype around Facebook targeting simply isn’t true. Self-proclaimed “gurus” even encourage people to use Facebook ads to steal competitors’ clients…That’s just not how it works. (Spoiler alert: that’s NOT possible).
Say for a minute we were talking about food instead of targeting.
If I tried a horrible mashed potato ice cream, I couldn’t generalize and say all potato dishes taste bad.
Chocolate ice cream might be great, but can I generalize and say any dish with chocolate tastes good? How about adding chocolate sauce to a hotdog. Dare to try?
The same applies to Facebook ads targeting; just because a case study about Lookalike audiences getting cheaper conversions exist, that doesn’t mean Lookalike will always convert better.
It all comes down to testing.
This said, there’s huge benefit to understanding how targeting works and clearing myths around it. That’s why we’re here.
Facebook ads targeting: the naked truth
Facebook can deliver ads to people based on their interests and behaviors. But that can be problematic as people can think Lookalike is always better or that targeting XXX all the time is more effective. And they end up making campaigns they assume are the better option when they haven’t tested other targeting possibilities. Here’s the problem with those one box wonder targeting strategies as I call them:
First, you might target the wrong audience for your niche. (Just like finding sausage in the middle of chocolate ice cream)
Second, you might avoid testing & targeting an audience that could have converted well out of assuming that a certain type of targeting always works better. (What if that “sausage” was actually made of chocolate, too?)
The proof is in the pudding. Well, translated to advertising language, I’d say: “the proof is in the testing.” The more we understand Facebook targeting and how it works, the better we can use it to our advantage and avoid costly mistakes.
This blog post will dispel the # most common Facebook ads targeting myths, and we’ll debunk them based on Facebook’s own documentation.
So, let’s start by shedding some light of knowledge on those myths.
#1 - Interest Targeting doesn’t equal targeting a page’s fans
Let me put an assumption to the test.
If I try “Lady Gaga” as an interest, I get the following:
Interest targeting gives me an audience of 110 million people, but when I visit the fan page, there’s only around 55 million fans.
On the other hand, MySQL at Facebook has 323,625 fans, but interest targeting returns an audience size smaller than the number of fans (unlike with Lady Gaga).
Then, there’s London's Lady Dinah's Cat Emporium cat cafe with 64,697 fans.
And people ask me: “I checked the cafe fan page, there are a lot of fans, but why can't I target her as an interest?”
That’s because interest targeting doesn’t equal targeting Page fans. (if it makes you feel better, here is a cat inviting herself to my drink at a cat cafe)
Then what’s interest target really?
Let’s go back to FB's own definition from their Facebook blueprint course below.
Image source: Facebook Blueprint
An older definition from Facebook included: “Interests may include things people share on their Timelines.”
On Facebook Business’ help center, they mention “what they've included in their profiles, as well as the Pages, groups and other things on Facebook they've connected with.”
Even keywords in a status update can play a role, as well as engaging with posts that contain such keywords.
Below’s a screenshot of an article on AdWeek in which someone from Facebook admitted to “showing related ads based on keywords used in status updates.”
Of course, we would never know the exact recipe of what constitutes an interest to target. And how Facebook algorithms define an interest targeting is something that continuously mutates.
A while ago, I posted horrible food pics on my profile for the sake of a laugh. A friend commented saying: ‘STOP SARAH’. FB confused her engagement with my post as the same as having an interest in the topic.
Later, she sent me a screenshot of the groups Facebook started recommending her.
Engaged shoppers aren’t people who love shopping
So many blogs with targeting tips say: If you’re running e-commerce ads, then target the ‘Engaged Shoppers’ behavior, because people who are known to purchase products from Facebook will see your ads.
So are ‘Engaged Shoppers’ people whose passion is shopping and buying products from Facebook?
The screenshot below gives us Facebook’s definition of an ‘Engaged Shopper’
Facebook terms it as: “ People who have clicked on the Call-to-Action button “Shop Now” in the past week”.
Based on this definition, let’s break down why just targeting this single audience might not magically give an advertiser more sales at a cheaper cost.
It includes people who clicked on the “Shop Now” button in the past week. So, no one who clicked on a Facebook post 17 days ago and bought a product will be a part of that audience.
Anyone who clicked on a post that didn’t have the “Shop Now” button will fall out of that audience even if they bought a product right after clicking on the post.
Just clicking a post with the button “Shop Now” doesn’t mean the person is a buyer.
A person might search for a book on Amazon, let’s say, and so Amazon will retarget them with an ad with the “Shop Now” button. In this case, they already bought a product for which they were looking, which isn’t the same as buying a product from a total stranger.
I’m not saying you shouldn’t target a precise category.
I’m just saying you shouldn’t blindly target that category and assume it’ll always convert better than not targeting that category or other targeting options. That’s like saying caramel sauce always makes food more delicious, which is not true if you’re eating mashed potato.
Lookalikes isn’t always the best option
An error I’ve seen in targeting orthodontists has to do with targeting 2100 orthodontists in France, for example, by using a 320,000 people lookalike audience.
Facebook lookalikes are set as people with common qualities in demographic or interests. But that can mean a broad set of multiple criteria; such as user behavior, profile data on age, gender, location, liked pages and clicked ads.
If you’re targeting a very tiny niche and orthodontics is the only similar factor, the stretch is too wide for targeting to make sense.
For narrow B2B niches, look alike might not be the best option.
Job Title targeting and Targeting CEOs
You could target job titles, for example.
Say something along the lines of “CEO” or so. People write their titles on their Facebook profile. The problem is a lot of people don’t add those to their accounts, also, and others play around with roles and add word play to their bios. Results thus aren’t 100% reliable. This isn’t LinkedIn where results are more exact.
For example there are 24,000 people on FB with the CEO job title who happen to work at Apple, Google or Microsoft.
The same can be true for people stating in their profile that they have a PhD from the school of life.
The above doesn’t mean you can’t get value from targeting some job titles. Just don’t imagine your ad is being served to the CEO of the world's most powerful companies.
People interested in Ferrari are rich
I’ll let the screenshot below do the talking.
There are 1.1 million kids on FB you can target with the ferrari interest.
An Interest targeting doesn’t equal a person who can afford a Ferrari. Facebook can be a place for daydreaming and dreaming costs nothing.
Which Targeting to use
Often, people ask me what strategy or audience targeting works best. And I say:
Test, test, test!
Marketers continuously talk about laser targeting ads, but you can’t get 100% of an audience to want a product or service.
Targeting is all about wondering, instead: To which audience is a potential client more likely to belong? And it comes down to testing multiple audiences, letting the user vote with their mouse and letting that number tell us which audiences worked better.
Leveraging the power algorithm for Targeting
The algorithm has a learning phase. Often, when we get enough conversions, FB can take care of the targeting.
Unless you have a very narrow niche, I recommend going broad.
It might convert well, it might not, but this all comes down to letting numbers speak.
How copy can help reach the right audience
Ad copy can help reach the right audience.
I once helped a financial adviser in the US who assisted parents paying for college.
At first, the ads attracted a lot of single mothers.
By changing the copy to mentioning that a couple with a job and a high salary can get a lower interest rate on a college loan for their kid, we started targeting a different audience.
Nothing else changed in terms of the targeting. We just changed the copy and attracted a different audience.