You’re running video ads in your FaceBook campaigns. Great! You should be. Right now, certain video ads are getting more attention and more engagement than static images with copy, for less.
So, how do you know which video ad in your huge test is best? The data feedback from FaceBook is dismal, but I may have a solution.
FaceBook shows you how many people saw your ad at the 3-Second, 10-Second, 25%, 50%, 75%, 90%, 95% and 100% intervals. As marketers, we are only concerned with 25%-100%, as the other metrics are very deceiving.
For example, if your ad is set to auto-play, then you will naturally see a huge percentage of people in the three and ten second columns. These data points are skewed because it may take three to ten seconds for someone to discover whether they want to watch you video, or not. They really didn’t watch your video with intent, so those two columns are useless for our purposes.
Over the years, I have developed a method to help gauge Performance and Engagement of a video ad, just by using the data in the 25 through 100 percent columns. This method is somewhat arbitrary and tends to confuse those who want concrete answers from their data. Think of this as a gauge. Just like the speedometer in your car, this equation is not exact, but is it better than what we currently have.
Let’s say we have:
Impressions 25% 50% 75% 90% 95% 100% 12,487 2,010 1,222 560 342 158 74
To calculate the Performance metric, we want to take the 25% number and divide it by the number of Impressions, expressed as a percentage:
2,010 / 12,487 = 16% Performance
This is the percentage of people who watched a meaningful part of your video ad. You would use this number as a rating to compare to other ads in the test or group. Easy.
The Engagement metric is a little more fuzzy, but it is easy to calculate. Simply add the number of views in the 25 through 100% columns, and arrive at a total:
2,010 + 1,222 + 560 + 342 + 158 + 74 = 4,366
Now, take 4,366 and divide it by the number of Impressions:
4,366 / 12,487 = 35 Engagement
Notice, I did not show this as a percentage, even though it could be. It is simply a rating. The higher the rating, the better. Also, because we added all the numbers, you could reach a number greater than 100.
Here’s why so many people give me that puzzled-dog look when I show them this metric. In FaceBook, when a person watches 25% of a video and then watches 50% of that video, they are taken out of the 25% column and placed into the 50% column.Why are we adding people who aren’t there? After all, only 74 people watched the video to completion, not 4,366. Those marketers are only interested in the number of people who completed the process. I am more interested in rating the popularity or overall interest the content provides the viewer. This equation rates the viewer’s interest, not the number of viewers.
In this example, 16% of the people who saw your video watched at least 25%. That’s the Performance rating. We also assigned an Engagement rating of 35 to the video. Some videos may show a Performance of 22% but an Engagement of 10. That would indicate the video had a better response at the 25% mark, but viewership fell off after that. The video started well, but interest waned, compared to others in the test.
Sure, you’d be able to figure that out just by looking at the numbers in each column. But, when you are looking at a sea of numbers from hundreds of ad campaigns, then this method makes it much easier to see the potential winners and losers. Remember, in all marketing, we are looking for trends. We are trying to predict the future, not analyze the past.