Background
We regularly produce square images to be shared on Facebook, and we are now moving towards making videos and sharing them on Facebook instead. We wanted to test the relative value of an image and a video conveying similar information on the same topic.
Experiment design
We prepared two Facebook posts, one with an image, the other a video. Both used the imagery of an Australian Medicare card in the colours of the US flag. It was about privatisation of Medicare.
Each post was given $100 in Facebook advertising and left to run to see how many conversions each post would produce.
Results
Version | Conversions | Reach |
Image | 148 | 25348 |
Video | 300 | 25017 |
Both posts had a similar reach from the same $100 advertising budget. For that reach, the video post had twice as many conversions.
Conclusion
This result is not definitive – some topics may be better suited to videos than images, and certain kinds of videos may be more or less effective at competing with an image. This 25-second video was designed to be viewable even without any sound – which is quite different to a video of a person speaking to camera.
It’s easy to imagine the effect partially wearing off as this kind of videos become less novel.
We plan to conduct further tests to confirm the effect, but it does appear that videos can produce a significant increase in conversions.
We are also interested in testing different types of videos – ones that involve spoken words as opposed to ones that can be experienced without sound, and videos of different lengths.