We’ve always touted the message “shorter videos are better”. They tend to get straight to the point and hold attention better than a drawn-out message. We’ve certainly noticed measured improvements when tightening up our own marketing videos. But are longer videos inherently less engaging than shorter videos?
Now that we have a wealth of video tracking data at our fingertips, we thought we’d put the old “shorter is better” theory to the test.
The Initial Test
We worked with one of our customers who has been testing two variations of the same video. The red line in the graph above represents the engagement of their 30-second video. The blue line is the engagement graph for a 90-second video. It’s expected that these two videos would have different engagement graphs, but what’s odd is that the first 30 seconds of both videos are identical!
Even though viewers were exposed to the same exact content, the drop-off rate of the 90-second video was much higher. By the end of the 30-second video there were about double the number of viewers than at the same point in the 90-second video.
A Wider Look
Our initial test got us thinking. Does this hold true on a larger scale? We decided to look at completion percentages across the thousands of videos we host.
We were right! Shorter videos are more engaging than longer videos, but there are some other unexpected trends. The average 30-second video was viewed 85% of the way through, while the average 2-minute video was viewed on average 50% of the way through. That is a very fast viewership drop-off. But what’s surprising is that the average viewing percentage stays quite consistent for 2-minute videos to 10-minute videos. That means the time spent watching is increasing almost proportionally with the video length during this period. Take a look:
What does this mean?
The data is quite clear, shorter videos are more engaging than longer videos. For videos 2 minutes and under, you should strive to make your content as short and punchy as possible to guarantee the highest engagement. If your video is 30 seconds or under, it’s very likely that most people will watch it all the way through.
If your message is more complex, be comfortable taking the time to explain it, but understand that half of your audience won’t make it to the end of the video. With this in mind, you would be well served to front-load your video with the most important parts of your message.
Understanding the relationship between video length and viewer engagement can help you ensure that your audience is getting the most out of your videos.
Interesting patterns. Could you provide more detailed data about this? Especially: What kinds of videos were these? Instructional? Commercial? Entertainment? Something else? And how many were there? And who were the target audiences?
this is great! i arbitrarily chose 1 min for all of my into videos… looks like it was a good call!
[...] out the full article titled Does length matter? It does for Video! blog comments powered by Disqus var disqus_url = [...]
[...] This post was mentioned on Twitter by Mark Nielsen, Web Startup Group, Alan Pinstein, kicauan, hrjn_rss and others. hrjn_rss said: Does length matter? It does for video: Comments http://url4.eu/195OI [...]
Very good to know! Thank you for the statistics!
The test was done using the exact same content
However the big unanswered questions are…
Was the opening on both videos identical ?
or
Was the same content just spread out over different time lines?
Long copy sales letters test better than short ones. The more you tell the more you sell. It has proven that long videos also have good conversion rates
So
How did the viewer know the length of the video he or she was watching?
Hi Rory,
Thanks for the comment! The first 30 seconds of each video was identical.
I agree, if that wasn’t the case, this whole exercise wouldn’t be nearly as interesting.
Great insight. Thanks for sharing this. We make videos for businesses and always urge them to go for short and sweet. This data helps us convince them.
[...] have been preaching for short demo reels for a while now. But this study adds actual data to the claim that people don’t watch long videos. What’s most interesting is [...]
I edit tv ads for a living. Many of those ads are reversioned ones from overseas companies. Most markets around the world tend to sell tv ad slots in 15/30sec chunks, but here in the UK you can buy 10/20/30sec slots etc.
So to maximise our spending, we always try to reduce a 30sec ad down to 20 seconds.
In the four+ years i have been doing this, i have not had a single ad that was not improved by removing 10 seconds of material.
Great post. As a producer of workplace training videos, the days of producing traditional 20 to 25 minute length training videos are long gone. Apart from being so time poor, training departments having to compete with other areas of their organisation for staff attention, and not to mention the attention spans of individuals… depending on the content, 3 to 4 minute videos are the most popular and is a formula we know works. We even produced a series of 60 second training videos that companies use for online training.