People are hung up on 'retention' when it comes to Youtube. You see it in the advertisements for people selling views all the time 'high retention views', but that is actually a smaller part of what determines a video's suggested video placement. Through the many conversations I have had with both Google/Youtube employees and product forum top contributors (who speak regularly to Youtube employees), the key to conquering the suggested video algorithm is not simply retention, but more importantly whether your video leads to a longer overall session time, regardless of whether that session continues on your channel or not.
Let me explain with an example.
Video A and video B are both 10 minutes long.
Video A is watched for 8 minutes on average.
Video B is watched for 4 minutes on average
After watching video A viewers typically, for whatever reason, log off of Youtube within an average of 4 minutes (after watching a couple more videos on your channel - a total of 12 minutes session time since hitting your channel, but all views being on your channel during that time)
After watching video B viewers typically, for whatever reason, continue to watch Youtube for an average of 15 minutes (after clicking onto someone else's video and watching videos on other channels - total of 19 minutes extra session time) before logging off of Youtube.
It makes little difference whether viewers watching these videos stay on your video's channel or click on a suggested video and continue to watch other channels. According to Youtube, Video B will be rewarded with higher suggested video placement due to longer overall Youtube session time being the single most important driving factor behind the Youtube suggested video algorithm placement.
So...the real key when it comes to trying to artificially generate views is to make sure that the IP address of those views continues to have a long but reasonable session time visiting random Youtube videos (or better yet, relevant Youtube videos) on other channels. Theoretically you could have the IP visit only a bunch of your own videos, but logic dictates that Youtube's algorithm might not think this was legit.
By the way, if anyone can provide views like that, I'd like to talk to that person.
Let me explain with an example.
Video A and video B are both 10 minutes long.
Video A is watched for 8 minutes on average.
Video B is watched for 4 minutes on average
After watching video A viewers typically, for whatever reason, log off of Youtube within an average of 4 minutes (after watching a couple more videos on your channel - a total of 12 minutes session time since hitting your channel, but all views being on your channel during that time)
After watching video B viewers typically, for whatever reason, continue to watch Youtube for an average of 15 minutes (after clicking onto someone else's video and watching videos on other channels - total of 19 minutes extra session time) before logging off of Youtube.
It makes little difference whether viewers watching these videos stay on your video's channel or click on a suggested video and continue to watch other channels. According to Youtube, Video B will be rewarded with higher suggested video placement due to longer overall Youtube session time being the single most important driving factor behind the Youtube suggested video algorithm placement.
So...the real key when it comes to trying to artificially generate views is to make sure that the IP address of those views continues to have a long but reasonable session time visiting random Youtube videos (or better yet, relevant Youtube videos) on other channels. Theoretically you could have the IP visit only a bunch of your own videos, but logic dictates that Youtube's algorithm might not think this was legit.
By the way, if anyone can provide views like that, I'd like to talk to that person.