Ladies and gentlemen, we found another Youtube Glitch. It’s not every day you can convince thousands of the top YouTubers on the platform to exploit the algorithm now; of course, if you’ve been on youtube for the last few days, you’ve probably noticed people spamming search terms in comment sections and that your front page is now filled with more polls than your average gentleman’s club so what enough happened to get us here well following the release of the youtube algorithm exploit video hundreds, and hundreds of YouTubers posted community polls.
The Youtube Glitch 2
Today we’ll be talking about the outcome and reach of this exploit. Equally, we will cover all of the new discoveries and even talk about the lovely response from youtube regarding the last article.
Now, of course, I have write many Viral articles before, like Why MKBHD is a Youtube Genius. Still, I’ve certainly never had a Article that influenced the front page of Reddit and broke the youtube homepage for thousands whilst also balancing a youtube feature into actually working precisely as intended.
The actual temporary fix to this exploit was actually to get creators to use the polls again. Ultimately the polls were being exploited more the fewer people use them, so if you flood the market with polls, suddenly every creator has 1 of the promotion instead of 5 channels just having 90 of it unknowingly.
Those lovely youtube creators have nerfed this exploit for me and actually made it so that this feature is working as intended so pat yourselves on the back.
But equally, the perks of having hundreds of creators actually make polls were that I was also able to use a much more comprehensive test pool to harness data from and using this data, my goodness we learned quite a few things about the youtube algorithm for a start youtube polls are more broken than I ever expected.
At the same time, we got one key aspect of the exploit entirely and utterly wrong.
So let’s begin with the most broken example of community polls. Allow me to introduce you to this channel that I recently discovered. It’s a channel with no name and 600 000 subscribers. The channel posts a community poll each day that manages to average between five and six million votes.
This is insanity. How on earth could a sister be so broken? Does the channel have five or six million subscribers? Does the channel even average five or six million views a video? No, should the channel be getting five and six million votes on a poll?
I’m gonna has to guess youtube and say that it probably shouldn’t. Oh my goodness, what a unique and creative example, so what did we get wrong? Well, it turns out all of you lovely people down in the comment section doing your lovely SEO bombardment. Guess what that’s pretty much partly meaningless polls are actually recommended to audiences based on channel tags and audience demographics, not any of the contents of the actual poll.
Let’s say you’re someone who likes to watch Minecraft, and I’m a Minecraft content creator. If you’ve liked my poll and you voted in it, then that poll is likely to be shown to other Minecraft people; however, if you started spamming search terms like Pewdiepie in the description.
It’s not going to be shown to the audience of Pewdiepie it can only organically grow that direction by having the chain of audiences all vote and like on the poll. However, this did actually have a very odd outcome because all of the SEO bombardment basically has driven up comments across the entire board of youtube.
The percentage of increase in comments on youtube community polls is terrifying. It’s a percentage increase that not even the stock price of GameStop can equate to, so basically, even though your spamming search terms has helped creators, it hasn’t done so in the way you intended.
It’s instead just driven up the engagement to stupid levels. Equally, something has happened that I can’t even explain in my own video.
The exploit actually was starting to get recommended to an audience who’d recently started watching Pewdiepie videos. This has never happened on my channel, and actually, this is quite a negative thing.
The reasoning is simple I don’t want Pewdiepie’s audience the lovely majestic sausages like yourselves who are watching this video are subscribed to me because you love breaking games, you love exploits, and you love destroying platforms most of you don’t watch Pewdiepie videos most of you watch content creators like the majestic arty game or call me Kevin.
I cannot actually explain at all why my video is getting recommended alongside Pewdiepie. He is not included in the tags, he is not included in the description, he is not included in the title I honestly have no idea, and now on to the next crazy interesting and broken thing spiff is now everywhere that’s right.
I’ve accidentally become your brand new youtube overlord this was not really done intentionally, and it’s entirely happening because people are responding to creator polls are now starting to get spammed my own videos this has meant that impressions on my channel have gone through the roof and equally so has my subscriber count.
It is terrifying I’ve gained over a hundred thousand subscribers in the last month. This isn’t normal. I might accidentally be hitting two million subscribers over the next few days, and I’m definitely not preparing for it now.
We did in the background just to see how broken these polls were was we set up a few dummy test accounts. These are brand new youtube accounts. They’ve seen no footage, and youtube has no data upon this technical bot account. Now what we did was we had these accounts watch an equal quantity of my videos.
Rt games videos call me Kevin’s videos basically creators who were similar to me they watched precisely 10 hours of content from each of these creators the only difference we made between these accounts was that the accounts all voted and like my community polls that were the only advantage I had as a creator.
Then what happened well the account started getting bombarded my videos over all of the other creators these are creators who I should basically have an equal footing to if not be actually slightly lower.
Because they have larger subscriber counts and audiences than I do, I’m getting a giant preferential treatment because the accounts voted in my polls. This is perfectly balanced. Oh goodness, it probably shouldn’t be working like that, and so if you’re a creator.
You’ve watched up to this point, and you’ve seen all of the other creators creating polls. You’re probably thinking, okay, well, if SEO spam doesn’t work, what is the perfect poll.
I’ve worked it out, ladies and gentlemen. The perfect community post to eat your channel into the stratosphere is a straightforward one you want to target your post for mobile users seeing as they makeup 90 of the people who actually respond to polls.
but most importantly, you want to make the poll as simple as it can possibly be a straightforward question with two choices allow me to demonstrate this glorious poll hereof are you alive its simple people who don’t know me will vote on it.
Anyone can get behind being alive or dead. In-jokes won’t perform well. More comprehensive topic discussions will generally memes will do quite good.
But suppose you want to entirely perfectly balance it. In that case, you want to keep it as simple as possible so what was the actual impact of my video and also the poll chaos we managed to hit the front page of Reddit with various memes being made about my poll as well as a few posts being made about the video itself.
But generally 90 of the focus was on the community poll and not the video explaining why it is broken equally twitter went into absolute chaos I was receiving hundreds and hundreds of DMS from content creators as well as being tagged in the posts of thousands of content creators showing off just how broken the polls features were thank you very much for all of the lovely support.