Exposing the YouTube Algorithm & How We're Using it Wrong

If you make anything for any online platform, you've likely heard of this all powerful entity known as "The Algorithm". The algorithm has the power to literally make or break what you're trying to build. Or does it? Go to YouTube and search for "YouTube Algorithm".

If you make anything for any online platform, and even if you don’t, you’ve likely heard of this all powerful entity known as “The Algorithm”. The bane of many creators’ existence, the algorithm has the power to literally make or break what you’re trying to build.

Or does it?

Please don’t get the wrong impression, I’m not trying to become the next “YouTube Guru”– but I am trying to combat a trend I’ve noticed. A trend that I think is doing a lot more harm than good. 

Go to YouTube and search for “YouTube Algorithm”. You’ll be bombarded with video after video trying to explain the secrets, hacks, and inner workings of this mysterious beast. Oh, and it’s constantly changing, so there’s a new wave of these videos every 3 months.

And I mean, who DOESN’T want to grow their audience? I sure do. Making things is tremendously rewarding, but honestly, it sucks to work really hard on something and have no one see it. 

IMG_3464.JPG

Algorithms aren’t good or bad, they just are. The YouTube algorithm has one goal: to keep you watching YouTube. Out of the nearly 500 hours of content uploaded every single minute of every single day, the algorithm tries to figure out what you’re most likely to click on to keep you watching. 

And don’t get me wrong, I’ve seen first hand how the algorithm giveth and the algorithm taketh away. The most popular video on my channel at the moment is an Apple Watch comparison. In the week after I made it, it was getting over 10,000 views per day, and then literally overnight dropped to a few hundred, and then drifted lower. Interest wanes over time, but this isn’t a natural drop-off: this is the algorithm picking up and then dumping a video. 

But regardless of stories like that, when it comes to explaining the algorithm, there’s the most important reality that all of these guides miss: We’re using the algorithm wrong. Yes, certain titles, tags, and descriptions make can make a difference, but there’s not reliable way to “hack” the algorithm or find its deep dark secrets. 

The algorithm is essentially a machine designed to think like people. In trying to “hack” the algorithm, we become people trying to think like a machine that’s trying to think like people.

And I think that’s crazy. 

The second most popular video on my channel is a Procreate tutorial for the iPad. This video sat around for six months, barely earning more than 100 views. Then one day I changed the thumbnail (literally nothing else, I just added a nicer, more interesting image), and it quickly jumped to 30,000 views and keeps growing. 

The algorithm probably didn’t know that one thumbnail was prettier than the other, but it probably DID notice that when that video popped up in search results after I made the change, more people were starting to click on it, which meant that the algorithm started showing it to more people who’s viewing habits were similar. 

Virtually everyone I know who’s had a video take off in some way says the same thing: “the videos you spend forever on get no views and the ones you put together without thinking get really popular”. 

That Procreate video was filmed in about 90 minutes as part of a 30 day upload challenge I was doing at the time.

And that Apple Watch video– the most popular video on my channel– I didn’t even intend to make it. I was actually in the middle of filming a different video when my Apple Watch got delivered and I thought it’d be fun to compare the new one against the old one. From the time I had the idea to the time it was uploaded was less than two hours. 

Now, this doesn’t mean that you shouldn’t put effort into your work. Both the videos I mentioned are still well made, and offer value to the viewer. But the main takeaway is that they’re very human. I wasn’t trying to predict what an algorithm would want or evaluate analytics and trends. I just made something I thought would be interesting, and put it together in a way where I would want to click on it if I saw it. 

And that’s the key. So here’s my proposition to any creators who are struggling with the algorithm: don’t. Don’t try and think like a machine that’s trying to think like you. Just be you. Make what you want and share it. 

For me, this idea is the best possible solution to the algorithm issue:

It gives credit to the smart people who made the algorithm in the first place. Instead of trying to “hack” it, just let it do what it was designed to do.

It takes the burden of off the people who make stuff and lets us make what we want.

And finally, it assumes the best intentions for all involved. Yes, I know that might be a bit naive, but at the end of the day, I’d rather assume the best in others than the worst. 

So what I hope you take a way from this video, is maybe a bit of confidence. Confidence to make the things you want to make, share them the best you can, and then move on to making the next thing. 

Go ahead and change thumbnails when you need to, make sure you’ve got good keywords, but don’t go overboard. Just make the thing. Then make the next thing.