NextAfter

How the aspect ratio of a Facebook video ad affects conversion

Experiment ID: #29631

NextAfter

Experiment Summary

Ended On: 03/31/2020

In this experiment, we had tested into using a talking-head style video ad for some audiences in our webinar promotions. But we wondered if we could actually get more registrations by using a square cropped video, which is Facebook’s recommended video aspect ratio.

The control was the full aspect ratio video ad. The treatment was the square aspect ratio. Both videos were exactly the same in every other regard, including the subtitles.

Research Question

Will a square cropped video ad increase webinar registrations?

Design

C: Full Aspect Ratio
T1: Square Crop

Results

 Treatment NameConv. RateRelative DifferenceConfidence
C: Full Aspect Ratio 0.39%
T1: Square Crop 0.08%-78.7% 99.4%

This experiment has a required sample size of 2,291 in order to be valid. Since the experiment had a total sample size of 6,788, and the level of confidence is above 95% the experiment results are valid.

Flux Metrics Affected

The Flux Metrics analyze the three primary metrics that affect revenue (traffic, conversion rate, and average gift). This experiment produced the following results:

    0% increase in traffic
× 78.7% decrease in conversion rate
× 0% increase in average gift

Key Learnings

The square cropped video ad led to a 78% decrease in webinar registrations.

In analyzing the results, I wondered if the square cropped video may have gotten more engagement, but fewer conversions purely based on how much the call-to-action gets pushed down when using the square cropping. Interestingly enough, the CPM cost we nearly identical and the engagement metrics for the square video were actually lower than the full aspect ratio video.

The square cropping seemed to fall short on every metric. 

One caveat that might affect the results is simply that this particular webinar promotion performed below our typical performance. But with such a significant decrease at such a high level of confidence, I’m not inclined to test this strategy any further.


Experiment Documented by Nathan Hill
Nathan Hill is Vice President, NextAfter Institute.

Question about experiment #29631

If you have any questions about this experiment or would like additional details not discussed above, please feel free to contact them directly.