“Big data is big nonsense” said Dennis Yu, who challenged attendants of BlitzU to draw real, actionable knowledge from their data analysis. In order to avoid data overload, marketers have to recognize when data is just noise versus when it becomes information.
Data, or analytics, should be measured in pairs or groups so you see the entire picture. Consider it algebra – an action on one side of the equation has to be balanced on the other. For example, aiming for a lower cost-per-conversion without considering the total-conversions might mislead you to false success if you bring the cpc down but also lose conversion volume.
If you only look at one side of the data when you analyze your Facebook ad performance – or anything, for that matter – you’re ignoring the effect your actions had on the other side of the pendulum.
Watching a swinging pendulum is fun… but at some point it gets old.
The burnout rate on Facebook is much higher than it is on Google, because there is intent when users search on Google. Burnout on Facebook occures when advertisers let an ad run until the frequency over-saturates their audience and the CTR falls through the floor.
So, how do you know when to stop running a Facebook ad?
First, you have to remove your emotion from that clever ad you wrote and think analytically.
Then, you start looking at the Frequency and watch for red flags. Alex Houg says that to identify ad burnout you analyze Frequency and CTR. Anything above a 2% Frequency might indicate burnout. To determine that, we need to balance our analysis by looking at the CTR.
If your CTR isn’t falling, then your audience is still engaging. However, if your CTR falls below 1% and your Frequency is higher than 2%, and your trendline is on a decreasing scale, then your ad is probably dead.