I recently produced a short promo video that performed surprisingly well in its first 5 weeks online. In fact, it’s out-performed any other video I’ve produced like it.
When a video performs well or badly, YouTube analytics is a handy and comprehensive tool to determine why.
You’ll find the analytic metrics in your YouTube ‘video manager’.
The metrics I tend to be interested in are, 1) traffic sources, and, 2) Audience Retention. There are a lot more metrics being tracked including views per country, viewer demographics, etc.
Traffic sources tells you if the video is being viewed from the YouTube watch page (organic search), as a suggested video by Youtube (someone watched something else and Youtube suggested yours to watch next) from an embedded website, etc.
Audience retention can be viewed as ‘absolute retention’ or ‘relative retention’.
‘Absolute’ tracks every minute your viewer watches your video. It shows you where they start dropping off and also tells you if they watch it more than once or watch certain parts over again (indicated when the retention percentage exceeds 100%).
‘Relative’ compares your video’s performance with other videos that YouTube deems similar to yours. Here you can see if your video is performing average, above average or below average.
As I mentioned in the book Run ‘n Gun Videography–The Lone Shooter’s Survival Guide, it is folly to expect that 100% of people watch your video all the way through. Even popular viral videos probably don’t achieve that. For one, there are those who click on it and click right off realising it was not what they were looking for. For two, they tend to click off when they perceive it is done if they do watch it all the way through, but may not stay on for any ‘end credits’. For three, people may stop watching when you’ve ‘sold’ them on whatever it was you were trying to do (in the case of business or fund-raising videos, product videos, etc.) –and that would be the purpose of the video in the first place.
This is not a metric, but it bears mentioning that you’ve all heard that videos should be ‘such and such’ a length due to the ‘short attention span’ of people. This is simply false. A video can be as long as it keeps the attention and interest of the intended audience period. Any other datum is simply the confession of video producers who produce crappy videos.
As an example, my most popular videos (the NX30 and X70 reviews) were both long videos. 14 minutes I believe for one and the other was even longer. Yet between them I think they’re well over 175,000 views with tons of engagement (comments, likes, shares, emails to me, etc.) and the audience retention is about 35% which I think is quite good. Look at it this way: 60,000 people watched the entirety of both videos.
But my recent promo for the Belvoir Castle shoot exceeded even that.
It appeared that Belvoir Castle uploaded the file I gave them directly to their servers, so that was not a source of the views counted by YouTube. The other site it was embedded on was Guns and Pegs (which was a YouTube link), but I doubted it was delivering that many views. So I had a look to see what was happening.
Turns out only 21% came from embedded websites. 43% came from the YouTube Watch Page and 1/3 of all views were because of YouTube suggesting the video when someone was watching something else similar.
Thus, nearly 3/4 of all views were the result of an organic search.
Next I looked at audience retention.
Here I found there was a very high retention rate (73%). Roughly 3/4 of all viewers watched the entire 4 minute video.
Relative audience retention (how the video performs compared with similar videos), showed it to be ‘above average’ through most of the video.
How did they find the video?
That is attributed to the relevant title, tagging and description I gave it. (mind you, this is for a targeted public–those who are interested in paying big bucks to shoot on a country estate, and to a lesser degree, those generally interested in the subject of shooting or Belvoir Castle).
Why did they watch as much of it as they did?
That can only be a measure of the quality of the video to get and keep their attention.
As you can see, using these analytics, one could go back and modify a video to improve it by seeing where attention drops off, evaluating what caused it and then remedying it. As for me, I usually just let it go if it seems to be doing its job.
There’s a chapter in the Run ‘n Gun Videography ebook that goes into greater detail on how to optimise videos for YouTube, and even then, there’s much more to it than I covered. But I did cover the essential basics based on my own experience.
Message is a big subject in that book and it might interest you to know that when I produced this particular video, the client had quite something else in mind. It took some fancy dancing to go ahead and produce it the way I did and then get them to watch it with a looming unalterable deadline facing us. After all, it was their interest I had in mind, not any desire to make myself look good. They loved it, so again, the point made in the book about the seniority of message was well proven as the video’s performance in the first 5 weeks has been very positive. You just have to understand and be able to clearly communicate the intended message and disregard ideas to the contrary.