Facebook Audience Insights is a great tool for advertising on Facebook but it can be dangerously misleading for research purposes. This is due to the way the segments are identified by audience insights, which will include a washed out prototype of your target not the precise segment you want to analyze.
To put the differences into context, let’s say our target brand is Jeep. When we insert Jeep as an interest selection on audience insights, the system will identify not just the people that interacted or liked the page of Jeep within a certain time frame, but anyone who since 2004, when Facebook was created has somehow connected or interacted with any “Jeep” related concept and has logged into Facebook in the last month. So for example if they liked a Jeep post in 2006 and logged in Facebook in the last month they will be included.
Facebook doesn’t disclose how these prototype and connections to the designated interest are created, but numbers prove how this platform is obviously created for advertising purposes. In this Jeep example for instance, facebook insights estimates the segment that is interested in Jeep to be composed of 25-30 Million people while the Jeep’s fan page counts less than 5 Million likes and all other Jeep related pages aggregated don’t top 500,000 other users. Furthermore the total number of yearly interactions on all these pages combined are done by fewer than 2.5 Million users. There is a difference of 20-25M users between an audience that is actively interested in Jeep versus what audience insights identifies as your target.
Audience insights actually allows the user to narrow down the target to the right segment, people connected to Jeep, but this is available only to Jeep page owner and managers.
Want to know more about this topic? Read the complete article here.