How do suggested videos work?

I haven't had much luck finding good information on this subject.

For the past 2 months, one of my videos has suddenly received a huge surge in views. The video is over a year old, and for most of its life it averaged 200-400 views per day. Then it started climbing, 1,000/day, 4,000/day, 10,000/day, and is now up to 20,000/day and still seems to be growing. Before all this happened, the video had about 100,000 views.

95% of this traffic is coming from the suggested videos sidebar in YouTube, and it appears to be spread around to a lot of videos. The largest referring video only accounts for 1% of it. I can only see 10 of these, and all of them are almost completely unrelated to the topic of my video.

So far, the traffic on my other videos has remained normal. So I'm trying to pin down the variables of this surge so I can try to reproduce it. The beginning of the surge roughly coincides with the time that I hooked up Adsense to my YouTube account, but I didn't monetize any videos for several weeks. The thumbnail of the video is a sexy woman in revealing clothing. While I'm sure that boosts clicks, I doubt this accounts for the fact that the video is appearing in the suggested video sidebar so much.

Sorry if this comes off like a brag. I am proud that this happened, but I have no idea why and would really love to repeat it for more videos. I'm just clueless how it happened.

I'm guessing nobody knows?

I've been thinking about treating it like SEO. Targeting certain high-traffic videos to piggyback with related keywords, or using backlink blasts. But I'd really like to hear from someone with experience or knowledge in the suggested video sidebar.

I suggest you look at the video reply whores and just copy their title/description/tags.

izSpK said:

I suggest you look at the video reply whores and just copy their title/description/tags.

Click to expand...

I've actually been paying close attention to this over the last few days, and give that a try yesterday. I'm not sure what to make of the results.

The case study video I used was uploaded just a few days ago and received several hundred thousand views. Let's call the uploader Bob. Bob is a partner and has about a dozen videos with 100k+ views.

I decided to try my hand at being a reply whore, to see if I could make it onto the suggested video sidebar for Bob's new video. Except I did something a bit more original than talking into a webcam, I created something humorous/entertaining that I thought people might enjoy.

My title and tags were very close to Bob's. Almost identical, with an additional word or two. The description repeated the name of Bob's video and included relevant keywords. I didn't do a great job at promoting my video (this is something I need to work on improving), so my video has a bit less than 1k views and a relatively healthy amount of likes.

I didn't succeed on making it to the suggested videos. I saw several other reply whores replying to Bob's video, all of them using similar titles/tags/descriptions, and they also failed. Actually, one person managed to get there for a short while, but he fell off within a few hours. His video also had less than 1k views and had more dislikes than likes, so I'm not sure what factored into his placement.

Most of the suggested videos in the sidebar of Bob's video are... more of Bob's videos! I wonder if this has anything to do with the fact that Bob is a partner, and if perhaps it was a bad video to use for this experiment.

EDIT: I realize that maybe targeting such a popular video is prone to failure. But if targeting a single video, there's no point in going after something small. So it's either that, or figure out how to get recommended for a large number of videos like my original video did.

Filename matters, try and tell me how it goes.

filename? why would that matter?

Yeah, I'm extremely skeptical that file name matters at all. Any evidence of this, even anecdotal?

Maybe he meant title.

astutenacute said:

Yeah, I'm extremely skeptical that file name matters at all. Any evidence of this, even anecdotal?

Maybe he meant title.

Click to expand...

I will try something this weekend and let you know, alright? Maybe even tomorrow on EpicMealTime's new video.

izSpK said:

I will try something this weekend and let you know, alright? Maybe even tomorrow on EpicMealTime's new video.

Click to expand...

That would be great, SPK. I can't wait to see your findings. Hooray for empirical data!

Just thought I'd let you know I've uploaded a video now with all optimized title/tags/etc, just boosting it with a little views and we'll see how it ends up.

izSpK said:

Just thought I'd let you know I've uploaded a video now with all optimized title/tags/etc, just boosting it with a little views and we'll see how it ends up.

Click to expand...

You're the best, SPK. Did you by chance experiment with the file name of the video?

astutenacute said:

You're the best, SPK. Did you by chance experiment with the file name of the video?

Click to expand...

As crazy as it sounds, yes I actually did. I doubt it very much though.

izSpK said:

Just thought I'd let you know I've uploaded a video now with all optimized title/tags/etc, just boosting it with a little views and we'll see how it ends up.

Click to expand...

Any results worth sharing with your experiments? I'm looking into this as a way to get a lot of real clicks on a video.

I've noticed that related videos have the extension &feature=related when clicked on the youtube site. I've also noticed

&feature=youtu.be (probably from their link shortner)
&feature=relfmu (related videos from same user)
&feature=branded (related videos from same brand i.e vevo 4od etc)

is it as simple as making a script so it plays your video with one of these extensions after watching another video? would this associate the two videos together?

hmmm very interesting

There is a great deal of randomness involved; think about it, if there is a certain process (like copying tags or whatever) that would lead to your video being suggested to thousands of viewers everytime you use it, someone is bound to notice it (if only by chance) and it would be exploited to death in no time. The only way for youtube to get around this is to randomize; so algorithm A is used 1% of the time, algorithm B is used 1% of the time,etc... which means you need more trials, testing with one video is not likely to prove or disapprove anything, if you notice a strong pattern that suggests something, test with at least 200 videos. Just my two cents.

Hey friends just wanted to know ....is Tube tumper & tube increaser still working in 2012, aftr the yt update???pls help me....if i use tube thumper for channel views will yt bann my act?????

Just to update you guys: it seems like the faster you upload a video with a "RE:" to the main video, the more your chances are to get put in the related videos sidebar.

Note to self: Start being a reply whore.

Adam Rock said:

Note to self: Start being a reply whore.

Click to expand...

Hey, ReplyGirl makes $10,000 monthly

From where? These videos doesnt have much views.