I did a small experiment on my channel where I published three videos and tried different strategies to see what makes people sub and like. To make the numbers comparable all videos was similar types of videos inside the same niche, they where also published during a similar time of the year (just a few days in between each video). They all had similar audience retention and about the same amount of views (around 7K views with the highest one beaning around 7.3K and the lowest 6.8K). One of them was shorter by a decent amount compared to the others.
Here are the results, comments are just the original comments, not counting replies.
I understand that this sample is to small for any serious research. There could be a difference in quality between the videos videos, maybe the one with the least subs where far worse than the others, maybe I just got lucky/unlucky with a video etc. I also wish all the video would have been about the same length.
But even considering these things I still found it interesting going through the numbers. Not counting the potential problems above there are a big difference in the number of likes and subs. Compare the number 0.0024 subs/view with 0.0101 subs/view, if you get 1 million views the difference is 2400 new subs compared to over 10K, that difference is huge!
That said even with this result I will probably never ask for subs/likes twice in a video again since I find it annoying and I'm pretty sure I got some dislikes from it and a comment or two about it. But consider the way audience retention go down during a video it may be worth asking for likes/subs earlier on in the video instead of at the end, if you can find a way to do that without annoying people. I wish I had made a video testing that combination.
Have you done any testing/experiments with your own channel and statistics? Have you found anything interesting? Please share!
Here are the results, comments are just the original comments, not counting replies.
Video 1 (shorter video) | Video 2 | Video 3 | |
Like and subscribe? | Did not ask them to like and subscribe. | I asked them to like and subscribe at the end. | Asked them twice to like and subscribe, once a few min into the video and once at the end. |
Audience retention | 52.6% | 45.6% | 47.9% |
Likes | 140 | 176 | 306 |
Likes/view | 0.021 | 0.026 | 0.042 |
Dislikes | 1 | 4 | 26 |
New subs | 16 | 51 | 78 |
Subs/view | 0.0024 | 0.0075 | 0.0101 |
Comments | 13 | 12 | 17 |
I understand that this sample is to small for any serious research. There could be a difference in quality between the videos videos, maybe the one with the least subs where far worse than the others, maybe I just got lucky/unlucky with a video etc. I also wish all the video would have been about the same length.
But even considering these things I still found it interesting going through the numbers. Not counting the potential problems above there are a big difference in the number of likes and subs. Compare the number 0.0024 subs/view with 0.0101 subs/view, if you get 1 million views the difference is 2400 new subs compared to over 10K, that difference is huge!
That said even with this result I will probably never ask for subs/likes twice in a video again since I find it annoying and I'm pretty sure I got some dislikes from it and a comment or two about it. But consider the way audience retention go down during a video it may be worth asking for likes/subs earlier on in the video instead of at the end, if you can find a way to do that without annoying people. I wish I had made a video testing that combination.
Have you done any testing/experiments with your own channel and statistics? Have you found anything interesting? Please share!
Last edited: