I’ve written before about the rise in popularity of video-based learning. Over the last few years, I’ve seen a rapid increase in the use of both professionally developed and informally produced video content. At first I was helped along by the wide availability of camcorders and flip-cams. But over the last two years in particular, the smartphone has become the video-recording tool of choice for many inside and outside of the L&D profession.
At the same time, as the concepts of social and informal learning have been adopted by organisations, the potential to grow the use of user-generated content (UGC) has arisen and excited those in L&D, keen to capture more of the knowledge retained in the heads of employees that would deliver greater value if more widely shared.
But motivating employees to create content is seen as a challenge that might hinder the use of UGC. If employees see this as a daunting prospect, then all the potential advantages will be lost. What can we do to realise our ambitions here and what are the implications for learning design and quality?
Motivating Employees to Share their Knowledge
I’ve recently explored these issues during the prototyping of a tool to capture user-generated video. We quickly came to the conclusion that short and sweet was the order of the day. Inviting an employee to create just a short video is an attractive proposition.
- Quite simply, it takes less time to create.
- More importantly, the shorter the video, the easier it is to get right. The longer the video, the greater the chance of making mistakes, such as losing your train of thought.
- There should be less need to edit. Longer videos often need editing, as the creator might well still plough on having made a mistake, especially if they are becoming more reluctant to have to start again from the beginning. It’s easier to start afresh, if the entire video is short.
- Short videos encourage the creator to simplify their messaging.
- Learners will appreciate the brevity of the content.
Before I answer the question about how short is short, let’s quickly consider the question about what do if the creator has much to say.
If your creator has a lot of content – and they’ve thought things through – then the advice is to encourage them to break their content up into a number of shorter videos. These can always be combined together later as a playlist, for example, but keeping them separate also makes the process of recording them all the more easier. The creator can tackle each chunk afresh and if they realise they need to amend something later on, they can just re-record the one shorter element. This approach also allows learners to pick and mix which bits they then watch, depending on their needs. I’ll return to some other tips for handling longer content later in this post.
Shorter Video Works
So how short is short? We mused about this for a while, before deciding to test out a maximum duration of 60 seconds. Yes, 60 seconds. One minute.
One minute appears to be an attractive proposition for someone needing to share some knowledge. You can actually say a lot in one minute and it definitely encourages brevity and provides a focus for the clip. Those with more to say are able to break down their message into shorter chunks, even thinking carefully about how best to structure their content, perhaps into a series of related steps or a number of quick hints and tips. This makes for much better learning content all round. Feedback from our prototype testers praised the speed and ease by which it was possible to create a short burst of video-based UGC.
I’ve also looked at data about video duration from outside the world of learning and development. Much analysis has been undertaken about the use of UGC on sites such as YouTube and Facebook. Both platforms are an enormous source of user-generated video content, so what can we learn from them?
|The average duration of the most popular videos on Facebook was just 81 seconds.
Engagement with a Facebook video falls quickly as the duration increases.
|Source: ReelSEO (2015)
The most popular videos being those with over 10,000 views.
Whereas the data from YouTube appears to suggest videos could be much longer, I’m drawn more to the data from Facebook. Why? Because Facebook videos share a characteristic that may only apply to certain YouTube videos. Facebook videos are designed to be shared. They are created in the hope they might “go viral”. And this is exactly what we want for UGC. We want that creator’s knowledge to be spread as wide as possible. So I’d be prepared to increase my maximum duration to 90 seconds on that basis.
Interestingly if you look on YouTube, its own search filters state that “short” videos are less than 4 minutes, whilst “long” videos are greater than 20 minutes. Four minutes is about the average length of all YouTube videos.
|The average duration of the most popular videos on YouTube was 14 minutes, 31 seconds.
The decrease in engagement with a YouTube video plateaus at 3 minutes, then falls quicker after 5 minutes.
|The most shared videos from news agencies on Facebook have a maximum length of 60 seconds.||Source: NewsWhip (2015)
Again, it would appear that shorter duration videos are shared more than longer.
|After 10 seconds, 20 per cent of viewers have stopped watching your video. Over 30 per cent will have dropped off after 30 seconds. At the one-minute mark, nearly half the viewers have moved on elsewhere. Only a third are still watching after two minutes.
|Source: Visible Measures (2010)
These abandonment rates remained the same regardless of the overall video length.
Clearly you have just 10 seconds to hook your audience in; and if your video is about 60 to 90 seconds in duration, a good proportion of the people will stick it through to the end.
|Abandonment occurs very early on, regardless of the duration of the video.
The longer the video, the more substantial the initial drop off rate.
|Source: Wistia (2012)
Wistia looked at the usage of many thousands of videos on its hosting sites. Again, regardless of length they saw a drop off early on. Their data showed that videos of around one minute performed the best, with 60% of the viewers still watching at the end.
For videos of 5 minutes duration, less than 40 per cent of viewers were still watching and within the first few seconds of the video starting, 15 per cent had stopped watching.
For videos of 10 minutes duration, by the end, only 30 per cent of viewers were still watching it and the initial abandonment drop off was also 30 per cent.
|After one minute of video, over 25 per cent of your content won’t be viewed.
|Source: Wistia (2012)
Wistia then looked at their data in another way. Where a video lasted a total of 30 seconds, on average, the viewers would watch over 80 per cent of it. With one minute videos, viewers would watch 75 per cent of it.
A five minute video saw viewers watch less than 60 percent of, whilst viewers would watch less than half of a 10 minute video.
|Product overview and demonstration videos are most popular when they last between 30 seconds and three minutes.
|Source: Animato (2015) and SAP (2013)
In different studies, the most popular video durations were between 30 and 60 seconds and between one and three minutes.
|The preferred duration of a formal learning video within a MOOC is six minutes.
|Source: edX (2013)
Based on an analysis of over 6.9 million video-watching sessions of between 1 and 40 minutes in duration and considering drop-off rates, it was found that the optimum duration was 6 minutes. More learners stuck with those videos than for other durations. Interestingly, many of the instructors thought that longer was better and many worked to an 18 minute duration, based on the average of TED Talks.
It’s all in the title
Interestingly, how you title and display a video title can also have an impact on its popularity and indirectly influence its suggested duration.
Remember, user-generated video content doesn’t just need to be viewed, it needs to be shared.
On Facebook, the most watched videos are those with titles under 40 characters in length. On YouTube, the titles can be between 41 and 70 characters [data from ReelSEO (2015)]. They also found that starting your video title with a question word, meant the video performed 23 per cent worse than those that didn’t.
Using a picture of a person as part of your video title suggests to the viewer that they about to watch something that is informal, fun, light and most likely short. If you use an image of a product, then browsing click-through rates are lower, unless the viewer is specifically looking for something about that product. And in a training context, the most viewed videos are those where the image is actually text-based, where the viewer assumes they will learn something from watching it [data from Advertising Age (2015)].
Top Tips for Creating Bite-size User-generated Video that Delivers Results
Taking on board this data (with some other results from the edX research), here are some tips to help you and your creators to get what you want from user-generated video-based content.
- Go for shorter durations – around a minute if at all possible.
- Start off with energy –speak fast and with enthusiasm in those important first seconds.
- Highlight the benefits of your video at the beginning.
- Start off with the important stuff. You might not have all your viewers towards the end.
- Include calls to action throughout at regular intervals. Don’t leave them all to the end.
- Break up longer pieces of content into a playlist of shorter clips.
- Choose your title carefully and keep it short.
- Give more complex content the time it deserves, but remember to front-load your videos with the most important information.
- Vary the pace – fast for easy stuff, slower for more complex and consider using graphics written or printed on cards to provide visual reinforcement and variety.
- Go for an informal filming setting.