Friday, May 18, 2012

What makes people subscribe to Youtube channels?

There have been various articles already written that attempt to explain why people unsubscribe from email lists, Facebook and Twitter but I don't think there has been any data about what makes people unsub from Youtube channels.

I personally have never had that many subscribers, so it's difficult for me to figure this out by looking at my own channel data. That said, I do have a few guesses based on what little data I have acquired from the lifetime of my oldest channel,


2,065 of the people who subscribed to my channel did so from my channel page.

After that, people subscribed after watching some of my videos; doing so by using the 'subscribe' button located on that video's watch page.


It seems clear there is no reason to believe that video view count is related to subscription numbers. My more popular videos don't acquire me any subscribers; for example, my Ultraman review has obtained 97,685 views as of this writing, but has earned my channel no subscribers.

I also lost some subscribers after people watched videos. However the number I lost is incredibly low; for the videos where people unsubbed, no more than 1 or 2 subscribers were lost.



I assume this was done by people who somehow accidentally subscribed to my channel or was an individual who was trying to make it seem like I was losing subscribers. Either way I think it is safe to say 0.000001% of the people who watched my videos unsubbed from my channel after watching a video.


The majority of my subscribers are people who live in countries where English is the primary language but I also have a number of folks who speak English as a second language.




You might wonder if I lost a significant amount of subscribers during the inactive account purges. I seem to have only lost 68 subscribers from these actions.

When I look through the dates of subscriber losses, I seem to lose 1-3 subscribers every other day. I'm unsure what to make of this, but it might be due to bots who sub to channels in order to build awareness of the channel and then unsub once they reach the limit of subscriptions they can have. Or the losses are from the people who try to do the "sub 4 sub" thing and unsub me when I don't sub back to them.

For comparison, here is the subscription data from the lifetime of my RPG Fanatic channel,






Very few unsubbed after watching a video, and the majority of the lost subscriptions seem to have been from the channel page or the account subscription page.

I have a great deal of difficulty trying to figure out how this data can be meaningful to me. If I was losing a lot of subscribers after making a certain video that would imply my activity is chasing people away..... but that isn't what I'm seeing. Even my controversial videos like the Star Wars : TOR rant led to people subscribing while many other videos I made did not result in any direct subscriptions.

The only thing I can think of is that perhaps I should make more videos like the ones which directly resulted in people subscribing to my channel. For example, a lot of people liked the Castlevania SOTN review enough to subscribe so perhaps I should do more Castlevania SOTN videos?

I also think the majority of people watching my videos probably aren't logged into Youtube, or perhaps they don't even have Youtube accounts? That would also explain why there is such a wide gap between the number of views and the number of people who subscribe to my channel.

That said, there does seem to be certain videos that have a "higher return" per view count; my MegaTen Spell names video has a very low number of views but has earned me more subscribers than the more popular videos. I suspect that if this video had tens of thousands of views I would have a lot more subscribers to my channel.

I do know one thing: analyzing this data allows me to figure out which of my older videos I should promote on places like reddit; the videos that encourage people to subscribe are the ones I should be using to introduce new viewers to my channel.

No comments:

Post a Comment