What is Facebook Audience Overlap?
Facebook Audience Overlap is a tool for optimizing and analyzing Facebook targeting that is as simple as it is effective. Basically, it allows us to select two or more previously created and saved audiences and evaluate the percentage of overlap of the two. In simple terms…
Think of your audiences as normal ensembles. Facebook Audience Overlap allows you to find out the number of users in common, therefore their percentage of overlap.
The types of comparable audiences
With Facebook Audience Overlap you can compare:
- Custom Audience
- Lookalike Audience
- Other previously saved Audience
The Audience comparison doesn’t need to be all of the same types, but of course, you must use a selection criterion such that the results have some specific relevance.
How do you use Facebook Audience Overlap?
First log into your Advertising account, select Tools> Audiences.
A list of all previously saved Audiences will appear. Select the ones you want to compare by ticking the checkboxes. At this point click on Actions and, from the drop-down menu, Show Audiences Overlap.
The Facebook Audience Overlap interface
As you can see, the tool uses a target audience and compares it with all the others you have selected. You can change the target audience by using the selector at the top right.
You can remove an audience by clicking on the X at the end of the row, while to add others you will have to click on Show Add Another Audience and select it from the textbox that will appear. By clicking, you will see a menu from which to choose, but if you have saved several audiences you will skim the results by starting to type the name of the audience you are looking for.
Why use Facebook Audience Overlap? Some practical ideas
In this paragraph, we will see some practical uses of Audience Overlap. In detail, I will explain how to use it to assess the need for a split test or engagement strategies, how to decide whether to target according to interest/behavior, how to create super-audiences and how to assign value to your fans. But let’s go in order …
Assess the need for a split test
In all the previous screenshots you have seen three audiences compared, it was not a random choice. These three audiences were obtained by creating Lookalike derived from the Custom Audience of Newsletter subscribers, visitors of the last 180 days, and fans of our Facebook page. I compared them to understand how sensible it was to test these three audiences for upcoming campaigns. If the percentage of overlap is very large, it may not be necessary to create a split test, especially if the budget reserved for the campaign is rather limited.
Assess the need for engagement strategies
If, on the other hand, I compare the Custom Audience, the overlap percentage gives me useful information about the composition of the groups of people who follow the various communication channels of the business. Is the audience made up of newsletter subscribers also following me on Facebook? If the answer is no and the desired outcome is different, I could – indeed, I should – consider setting up a strategy.
Deciding whether to target according to a particular interest
Thanks to the evolution of the tracking tools made available by Facebook, targeting by interests or behaviors is less and less necessary. However, in some cases, they can still be quite useful. Just to give an example: especially at the beginning, you will create Lookalikes starting from sparsely populated Custom Audience, thus sacrificing affinity. In these cases, it is worth “tweaking” the audience created by filtering again using interests. But which interests to choose?
To determine which ones are the best, by affinity, you could save the reference Custom Audience – for example your Facebook fans – and then save an audience for each specific interest you want to test. The ideal situation would be one in which your fans make up a subset of the isolated audience with interest. This would mean that your audience is very close to the interest test.
Conversely, if none of your fans are involved, you would have the certainty of being able to exclude that particular interest from the target audience you are building. Of course, you will always find yourself having to evaluate intermediate situations.
Please note. For a particularly small Custom Audience, you may not reach the statistical significance threshold.
With the same method illustrated in the previous paragraph, Jon suggests adding interests/behaviors to the saved audience trying to increase the number of users involved and, at the same time, keeping the affinity intact. This will allow you to create very large audiences without sacrificing ad performance.
Understanding the value of your fans
Another very interesting idea. How Much Are Your Facebook Fans Worth? In other words, how many of your fans are also your customers? You can answer this question with Audience Overlap very easily. Create a Custom Audience of your fans and another that includes all your customers. The intersection between these two sets will tell you how many of your fans are also your customers. Obviously, this parameter will have to be the basis for strategic reflection.
Before Facebook Audience Overlap
Audience overlap analysis was possible even before Facebook Audience Overlap. So while it is true that the tool is new, the practice really is not – or at least not completely.
Overlap using Targeting
Using the targeting we can create an audience of users who have expressed interest, for example, to Social Media Examiner And for AdWeek. The resulting audience size is nothing more than the overlap of the two audiences. Of course, in case you are managing a page with a fanbase large enough to appear among the interests, you can also use this method to compare your audience with others.
Another way to get a similar answer is to select an interest – let’s use Social Media Examiner again as an example – and then filter out the people who have NOT expressed another interest – again, AdWeek. In this way, instead of overlapping, I can evaluate the number of users belonging exclusively to set A.
Clearly, by selecting a previously saved Audience as a user base for the ad group, we can find the intersection of unique users using the same operating methods. We can also use both methods to create targeted audiences.
As you can see, even with interests we can obtain similar results, however, compared to Facebook Audience Overlap we have two main limitations :
- Certainly less immediate procedure
- Inability to compare two or more previously saved audiences with each other
Overlap using Audience Insight
Using Audience Insight you can analyze a Custom Audience, for example, your site visitors, and by clicking on the Page Likes tab, you will find the number of users who have browsed your website and who at the same time are your fans on Facebook. You will also find the degree of affinity between the two audiences (visitors and fans). However, to access this type of data, you need to be able to rely on rather large audiences. To give you an example, the Custom Audience of Coinè visitors is made up of 1,300 users and Audience Insight is not able to generate a report on their interests. This limit exceeds the normal barrier for audiences of less than 1,000 units to protect the privacy of users, a limitation, the latter, which also applies to Audience Overlap.
Also in this case, therefore, we have some disadvantages compared to Facebook Audience Overlap, in detail:
- We cannot evaluate the intersection of two or more Custom Audience
- We cannot take a Lookalike Audience into consideration
- Although the audience is made up of more than a thousand Audience Insight units may not have enough data to generate the reports
Defects of Facebook Audience Overlap
As mentioned earlier, the most obvious flaw consists in the impossibility of comparing an Audience below 1,000 units, so if you are starting you will have to wait sometime before you can benefit from this tool. Facebook motivates this limitation as follows: “To protect the privacy of people on Facebook, we don’t share data for audiences of fewer than 1,000 people.”, is the same message that appears on Audience Insight and is all in all reasonable.
If we want, then, it would certainly have been more convenient and faster to be able to directly select the audience of a fan page, an interest, or behavior rather than having to save an audience every time. However, more than a real limitation, we are talking about an inconvenience that will probably be solved with some future updates.
To sum up this long analysis we can say that Audience Overlap is certainly a useful tool. Improved, of course, but very well-conceived already in its first iteration. Certainly, it cannot, must not, and does not want to replace all other methods of comparison. On closer inspection, in fact, both the comparison using interests and the one obtained thanks to Audience Insight, even without giving us all the information obtainable thanks to Audience Overlap, offer others that, on the other hand, are not available thanks to the newcomer. In this sense, Audience Overlap is another brick with which to build the audience of our campaigns with ever more precision.
For website maintenance service contact us.