Awhile back, we pointed out a video that obviously paid for its views, or found some sort of loophole in the system that had trumped up the view count. These bought views (spam), either by bots or other services, obviously don't represent real engagement, but it's done to show that it can be done, or to get your video to rank better than it normally would (the quintessential definition of Black Hat SEO), or hopefully leading to legitimate engagement. These videos aren't always easily spotted, but there are some clues that should cause you to question a video or channel's engagement, and here are a few of them.
Signs A Video's Engagement is Spammy
The crazy, rapid rise of a video in the YouTube ranks does not, in of itself, mean it's got something shady giving it a boost. But after a certain amount of views, you should expect a certain amount of comments, likes, shares, and favorites that show legitimate interaction. But not surprisingly, you need to watch out for those engagement scores as well.
Lots of Views, Little-To-No Comments
It doesn't even have to be millions of views. Maybe over 10,000. There is a certain point, likely in the hundreds but certainly in the thousands, where a video should be getting some sort of engagement. If you see something with tens of thousands of views and no comments, and the comments are not turned off, then I'd raise a red flag on those. Even if the comments are in single digits, and the views are in the tends of thousands, it doesn't compute really. The chances that you've found 10,000 people to watch your video and none of those people are the type to comment are pretty low.
The same kind of thing can be said for likes and any other feedback. Maybe this is a bad example, but with a Freddie Wong's "Guns - With Smosh" video that has 11 million views, we see a ratio of about 1 like for every 200 views and 1 comment for every 700. Now, should it be this exact ratio? Not exactly to the letter of the law. It is a small sample size to look at one video, but it gives you the kind of ballpark numbers you should be looking for. Rebecca Black's reviled "Friday" video still has a 200:1 like ratio (and about a 48:1 dislike ratio) but a 70:1 comment ratio, and probably because the video struck an emotion in many to tell her just how much she ruined everyone's life posting it.
Lots Of Comments That Praise A Video In A Repetitious, Robotic Way
Or we can call it what it is, spam. Now, spam can work in two ways:
- It's either spam-bombing a legitimately popular video in order to get people to try to click on other channels
- Or it's spam that makes a video look like everyone's talking about it.
Awhile back, we got an e-mail from someone about this video on YouTube possibly being black hat. Let's take this one as an example, shall we?
The comments on this video are almost all during the same time, and nearly all of them overpraise the content. There are a lot of weird, stilted phrases. The person who tipped us to this thought it might be a sign that cheap labor was hired to farm out lots of comments. And just look at the video...there is really no reason why something like this would garner many views or attention.
It's also weird how if you go to the second page of comments and then click on the 1st page again, how some of those comments look like they've been flagged as spam and removed, but they still show up when you first arrive to the video.
By the way, I did check this channel's other videos
Unusually High First Referrals From Questionable Sources
When you suspect that a video is using black hat SEO, you can click on their YouTube stats next to the view count in most cases. Again, this is something that can be turned off, so not always. But back when we talked about this video, one of the strangest parts of the analytics under "Significant Discovery Events" was that, out of a million or so views, over 600,000 had its first view from a mobile device.
Now, mobile devices are popular, don't get me wrong. Go to virtually any video on YouTube and the mobile device will almost always be a significant discovery event with lots of views attributed to it. But fully 60% of the views, when there are no other significant events attached to it?
Reasonably popular videos have several events that work to make the video popular. It was linked from a popular blog, or it picked up steam from being a Related Video to something that was already raking in views. This particular video had 600,000 first views from a mobile device, and all the other significant events were in the low tens and single digits.
But that can work in another way: if you see 600,000 referrals from "Bob's Blog" and "Bob's Blog" doesn't have nearly the kind of traffic that would generate that kind of view count, that's probably shady.
Use Common Sense
Videos that are truly good and have a lot of buzz outside of YouTube are getting legitimate views. There are probably some videos out there that have used black hat spamming to get to that point, but in the end, truly interesting videos don't really need those methods. Sometimes, it's as "simple" as getting a notable blog to share your video and it takes off from there.
But if you see a video that has nothing funny, or embarrassing, or unique, or interesting whatsoever, or at least interpreted as having those things, and it's climbing the charts or making its way onto the YouTube front page, then red flags should be raised. Many of the changes YouTube has made in in its history have tried to combat these shady methods, and it's part of the reason why they keep changing seemingly every month.
Does YouTube Black Hat Spamming Work?
There are numerous ways, though, that black hat SEO can, have, and will be cashing in. It might not even require a couple of videos that find their way into viral status. Many of these people open up hundreds of YouTube accounts and post videos that way, where no one particular video does super well but with a ton of videos all collecting pennies here and there, they add up. And if one does happen to go viral, they can capitalize on it by re-posting it several times. They change URLs, keywords, video length, tags, descriptions, and so forth so that YouTube doesn't recognize those videos as duplicates. So you may not be able to spot black hat SEO on a particular video, but if you do a lot of digging you might be able to find videos that are either published more than once or outright stolen from other channels.
I have a feeling that's where black hat SEO gets most of its success. Videos we have all seen, know, and love/hate are not likely the product of black hat. They have to stay a little bit more under the radar to keep doing what they're doing. But hey, if one of their own videos happens to find its way into the public consciousness, it's probably deserving of it because it's good content.
My conclusion is that black hat only truly works with massive scale: throw as many videos under a ton of cloaked accounts at YouTube as possible and see what sticks. There are programs that do it and require little to no work. I feel like that's YouTube's problem to fix, that with all this incredible growth comes the idea that it's impossible to police effectively. The stupid ones will get caught, and the smart ones will always find a way.
Don't Be a Stupid SPAMMER!
Why not? Even though someone may be able to fool clients into thinking that they were able to generate oodles of views and possibly game the system such that the video performs well in search, it really doesn't matter when the outcome is as it was for Daniel Cohen.