When Youtube first launched in 2005, it was easy to rack up really high view counts; all you had to do was go to your video and reload the page over and over and over. The site counted each page view as a video view so if you wanted to have an impressive (but fake) view count, you could install and run a plugin that would “reload and refresh” the page every 5 or 10 seconds. Youtube’s engineer’s caught on to that trick pretty quick and soon the implemented the site’s mysterious 301 rule. You’ve probably noticed that sometimes Youtube videos get “stuck” at 301 views, right? Well there’s a technical explanation for that phenomenon. The first 300 “views” were actually just page views. So if you refreshed your browser over and over, you could generate a bunch of free (i.e. fake) views. But once the counter hit 300, Youtube would start scrutinizing the source of your views. If too many views came from the same IP, the counter would freeze until the TRUE view count caught up to the page view count. And if a video was actually going viral, and if multiple views were registered at the exact moment the counter hit 300, the view count would freeze at 301, 302, 303, etc views.
Confused? If you are, don’t worry about it because last year Youtube refined their code and did away with the 301 system. Now that the changes have been fully implemented and tweaked, I decided to run a little experiment. On June 23rd, 2016, I uploaded a very boring video to Youtube. I gave it a simple title that probably wouldn’t come up in anyone’s search results. Here it is: –
I let the video sit for a few days so the file could spread across Youtube’s servers. (When you upload a new video, copies are sent to severs around the world. Once the process is complete, the view count starts showing views from other regions). Then I used a Firefox plugin to reload my video’s page over and over.
When I started this process my video had 4 views. I set my browser to reload the page every 10 seconds. Every time the page reloaded, a new view was added to the counter. After 50 reloads, the counter froze at 54. I let the plugin keep going got another 10 minutes but the counter was stuck. – –
I shut off the plugin and closed my browser. I then left the video alone for a few days to see what would happen. The next time I checked it, the view count has actually gone down. That means Youtube automatically checked the validity of ALL of my views. It found too many from the same IP address and so those views were removed. – –
I can understand why Youtube made this change but frankly I think they went a little overboard. Where’s the harm in letting people generate a few free views? A brand new video with only 2 or 3 views looks a little pathetic. So I used to reload my videos 20 or 30 times just to rack up some views before I sent it to anyone. Now that I think about it, these new changes will probably inspire more people to buy fake youtube views. And it’s those folks that are really damaging the integrity of the site.
If you poke around the microjob website Fiverr.com you’ll discover that you can have thousands and thousands of “views” delivered to the youtube video of your choosing for just a few bucks. But don’t get excited; you can’t just buy youtube super stardom. When you order one of these gigs you’re actually buying hits that are disguised to look like views. So 15,000 real people will NOT be watching your video. Your view count is simply going to get inflated by 15,000. Some of these bogus views are generated by bots but in many cases hackers will secretly embed a video on a popular web page. Every time someone accesses that page the video automatically plays….albeit silently and invisibly.
Fake views are really starting to have a big impact on the way youtube works. Google’s adsense program is getting scammed out of lots of money and lame, wannabe Internet stars are wrecking the site’s integrity by purchasing phoney viral success. It seems like the folks at youtube are finally fed up and they’re now cracking down on users who buy fake views. A few weeks ago youtube actually removed more than 2 billion fake views from lots of big name companies and they even shut down the channels of some of the more egregious offenders.
It’s easy for youtube’s engineers to spot fake views since they can see data that the general public doesn’t have access to. But what about the rest of us? A lot of video contests use youtube views as part of an entry’s score. How can you tell if a contestant is cheating by ordering fake views?
Unfortunately you can’t….at least not for sure. But these scammers leave a lot of clues so it’s easy to spot views that are probably fake. Let’s take a look at a youtube video from a channel that hosts close to 300 self-help/get rich quick videos. It’s is run by some dude who wants to teach you how to apply martial arts lessons to the world of real estate….or something like that. Yes I’m being serious.
WOWEE! 2.1 million youtube views! That guy is gonna be the next PSY! Or maybe not. Maybe he’s a con artist who bought millions of fake youtube views because he wants to make it look like he’s popular so that suckers buy his book or send him donations. I’m not saying that’s the case….I’m just saying MAYBE that’s what’s happening here. Why do I think that some of those 2.1 million views might be fake? Here’s a quick run down of all the red flags I see in this screenshot:
1. The view count is suspiciously high: You should always be suspicious when you come across a video that has an inexplicably high view count. Regular youtube users understand what types of videos get a lot of views and this particular video just doesn’t feel like it should be popular. Would 2.1 million people really watch a 9 minute long get-rich-quick video?
2. The channel has a relatively low number of subscribers: 10,994 subscribers might sound like a lot but that’s a ridiculously small number for anyone who is getting hundreds of thousands or even millions of views on most of their videos. Here, check out the stats for a GOOD youtube channel; The Final Cut King. The FCK (heh) gets about a million views for each of his videos. But unlike the real estate guy he has more than 220,000 subscribers. Those are numbers that actually make sense.
3. The video only has 2,254 likes: Guess what? You can buy likes and comments on fiverr too. But those are way more expensive than views. So people who buy fake views sometimes don’t even bother ordering fake likes and comments. Consequently the stats for those videos are totally lopsided. It’s just not realistic that only 1/10th of 1% of the people who watch a youtube video will click the thumbs up or thumbs down button.
4. The video only has 20 dislikes: Few scammers are going to be smart enough to order DISLIKES. Two million views and only 20 dislikes just scream fraud to me. If millions of people had actually watched this video a few thousand of them would have clicked the thumbs-down button. That’s just how things work on youtube.
4. The video only has 489 comments: This number also doesn’t jive with the view count. Once again, let’s use the Final Cut King as an example. Here’s one of his videos that has received 2.5 million views. It has close to 6,000 comments (and over 30,000 likes/dislikes.) Comments are REALLY hard to fake since each comment needs to come from a different account and the comments need to be spread out over days and weeks to look realistic.
So the basic stats here seem pretty fishy. Let’s drill down a little deeper into this video’s data to see what else we can discover. If you look under a video’s view count you’ll see some symbols. If you click that symbol that looks like a bar graph a bunch extra stats will pop up.
If you click on the stats button for the get-rich-quick real estate video, this is what you’ll see. The arrow indicates the one gigantic red flag I spotted.
Now we’re getting into the really good stuff. This data shows us where all of those 2.1 million views actually came from. It used to be that if someone ordered fake views you could tell right away because the referrer sites would all be pages like buycheapviews.com or whatever. But the view-sellers have gotten smarter and they’ve figured out ways to mask the origin of their bogus views. And the number one way they do that is by tricking youtube into thinking the views came from a “mobile device.” Does it really seem possible that 70% of the 2.1 million people who watched this video watched it on their their iPhone?? That’s just ridiculous. And the facebook and Twitter numbers are almost certainly fake too. View sellers like to embed their videos on those sites because it seems natural that a lot of views would come from popular social media sites. If this video does have fake views, the video was probably also secretly embedded on all the other sites listed. After all, how the hell can you get thousands of views from a youtube video that’s been posted on iTunes or LastFM!? The data that you don’t see here is also suspicious. If this video really had gone viral you’d see referrals from specific blog posts or news stories that featured the video. And if the channel has any real subscribers, you would also see lots of views listed under “referral from a subscriber module.”
Let’s take a look at the traffic stats for a different video. In 2011 I was in a video contest where the finalists were determined by youtube view counts. This contest turned into a total clusterf*ck and people cheated like crazy. I didn’t know what was going on but after doing some research I quickly discovered that anyone could simply order youtube likes, comments, subscribers and views. These are the stats from one of the cheaters who made the finals.
Check out the item that’s 2nd to last in that list. 2/3rds of this video’s views came from the website Viralzoo.com. What the heck is viralzoo.com?? If you ever see a weird website listed in the stats, check it out. In this case, Viralzoo.com is just a bogus page that’s used as a front for view-fakers.
There’s another piece of very important data in this image. Look closely at that blue graph….
This graph shows how quickly the video got its views. You can see that the view count suddenly skyrocketed after the video was posted to those referral sites. All those referral happened BEFORE the video became popular and zero new referrals happened after it got all those views. That means that nobody liked the video enough to share it ANYWHERE on the entire Internet. And that big plateau means that the view count has basically remained the same for a year and a half. So from this data we can tell that the video was posted to a few different websites all at once. Then the video gained almost all of it’s 49,000 views over a few days but then suddenly the video somehow went un-viral and pretty much nobody every watched it again.
So….you can find out some pretty amazing stuff if you check the public stats data of a youtube video. Unfortunately youtube lets userrs turn that option off and most people who buy fake views are smart enough to hide their referrals. But turning off the public stats is a pretty big red flag in and of itself. This is what you’ll see if a user wants to hide their referral data:
If you’re ever in a video contest where youtube views matter you should tell the judges that they should ask all the contestants to leave their Public Statistics data ON. There is no legitimate reason for a contestant to hide the source of their views.
Here are a few other tips for spotting fake youtube views:
1. Check the youtube channel of the person who posted the video. Many youtube users will post links to their facebook pages, blogs or twitter accounts on their youtube channel. You can check these sites to see if the person is genuinely popular. If a video’s stats say that 50,000 views came from twitter then the creator of that video should conceivably have a pretty massive twitter following.
2. Keep an eye on the number of likes and comments. If a video gains 30,000 views in one day but receives zero new comments or just a few likes you know something shady is going on. Popular videos always get a steady stream of likes and comments. If a “popular” video isn’t getting any comments it means that real humans aren’t actually watching it.
3. Read through the comments and look for anything hinky. A lot of the people who sell views, likes and comments live in foreign countries like India. So fake comments sometimes don’t make sense or are done in broken english.
4. Check out the channels of some of the people who have left comments. If you look at their activity feed you might notice that they have been doing tons and tons of really spammy stuff like commenting on dozens of get-rich-quick videos or liking every music video created by one random aspiring hip hop artist.
And finally, I’ve saved my best tip for last: Just stay away from video contests where youtube view counts or likes help determine who the winners are. Like I said, there is no way anyone outside of youtube can be 100% sure that a video has fake views so cheaters will always win view count races. Sure, you could try and contact the judges and tell them how easy it is to buy youtube views but they probably won’t give a damn. If they actually cared about running a fair video contest they wouldn’t have run the thing on youtube in the first place.