From Data to Insights on YouTube
Into the Viya Verse Thumbnail
In this post we will get to my main goal with my YouTube channel and how I'm starting to turn data into insights to help me get there.
During my trip to the SAS headquarter in North Carolina I had an illuminating discussion with Brendan Bailey about the videos I am currently creating. We got to talk about what my actual goal is and how to go about achieving it better.
To get the core of my goal with creating videos we will first have to define what it isn't. Gaining subscribers is an amazing feeling (even though ~90% of viewers are not subscribed) , it is a number that has already kind of lost its meaning to me. This metric is not relevant to me, as I have no ambition of ever earning money with my channel - otherwise the first target on my list would be to get to that magic number of 500/1000 subscribers to become a YouTube Partner.
Next on the list, and related to subscribers, is the number of views per video. Well if I cared about that I clearly shouldn't be doing videos about Custom Steps because they do no generate any interest at all. Again views are of course great to get, if nobody was watching, then it would be kind of a downer. But I'm comfortable with the level of views I am at. So again not my target.
Number of likes and comments is a weird one to me. Comments are good, but even with such a small channel I already get spam comments at least once per video - so just getting more of them isn't really of interest to me. Likes are again cool, but they do not really provide any valuable feedback and I'm not going to start shouting the YouTube famous: Like, Comment, Subscribe and Hit the Bell.
These four metrics are the once that anybody can look up on a channel. You can just go to tracking websites or my channel directly and see how these metrics look like. Two metrics are not really visible from the outside though, and this is where things get interesting to me.
The first of the two is repeat viewers. That means people that come back and watch additional videos after some time. Of course this metric is correlated with subscribers, but taking a look at big YouTube channel you can see that this metric isn't causal in its nature. I'm assuming this is also why YouTube started to introduce the Bell. People started to subscribe to so many channels that it became to many notifications so a second layer was introduced. I can not wait for the inevitable third level that will have to be introduced - Button > Bell > Gong? Why do I care about this? Well its because it shows me that my content provides value to the viewers so that they return when new videos are published.
Now finally we get to the metric that is my goal Audience Retention. Audience Retention measures what percentage of your audience watches your video for how long. So at second zero you have 100% and then its start to drop of over time. As this statistic isn’t public I found a lot of claims around 50-70% being the verage retention rate on YouTube, but then found this refreshingly honest Reddit thread that seems to suggest it might hover more at around 30% [source]. I care about it, because I believe that if you take the time out of your day to watch the whole video it was valuable for you and you got what I promised to you by my thumbnail, title and intro - and I certainly do not want to produce clickbait.
For this blog post we are going to compare the metric of two of my entries in my Into the Viya Verse series - episode 8 and episode 9. They are of course not 100% comparable but they are the closest things I have in terms of an apple to apple comparison. Both videos are around 10 minutes long and they target the same audience of people. And episode 9 is the first one after my conversation with Brendan Bailey.
Let us start of with episode 8, by taking a look at its Audience Retention graph down below. Let me quickly explain to things about this graph:
the dotted vertical lines in the graph represent YouTube video chapter marks - aka timestamps
There are ups and downs in the graph, that is mostly related to people jumping around in the video or going back and re-watching a certain section.
Retention-graph for Into the Viya Verse Episode 8
We see a huge drop off in viewer ship just in the first 30 seconds - I lost around 45% of people that initially clicked on the video. And only 25% of people watched the video until the end. That is a crushing statistic, so only one in four people for it was worth their time to watch this video all the way through.
Before I talk about what I changed for episode 9 let us take a look at its graph:
Retention-graph for Into the Viya Verse Episode 9
The average watch time jumped from 3:20 to 5:02 that is a great increase. Also at the end 39% people were still watching. Both are these are great gains!
So what changed? What Brendan Bailey shared with me was all about how to make the first 5-20 seconds count of my video to create a more engaging intro. Talking with experts is the way to gain insights from your data, and then use data to confirm them. What was changed:
Repeat the title message as early as possible
Give a reason why it is worth watching until the end
Show a bit more of my excitement
If you have the time please watch the first 20 second of each video and compare them for yourself - which are you more likely to watch all they way through? Well actually please watch the whole thing, I do not want to ruin my stats.
This was a great success - If I wanted to do a clickbait title it could have been: 3 Tips to boost your audience retention by 80% - but clearly this was just my first experiment, lets see how it plays out overtime (sample bias is a thing). And also audiences change over time so I will have to keep an eye on that. But this again shows what happens if you combine data with expertise to generate insights. And then you will have to monitor how trends shift. Thank you very much Brendan Bailey for sharing your knowledge with me it was a pleasure to get to connect in person with you!
