Have you every thought of watching a particular YouTube video and have it pop up on your homepage the next time you open it?

Watching YouTube videos is one of my favorite past times and I am sure you can relate too. People spend hours watching videos on YouTube daily. From how to videos to cute puppy videos and everything in between, YouTube is an integral part of our life.

But how does YouTube know what videos to show on your homepage? Let’s find out!

What is the YouTube Algorithm?

For those of you who may not know, YouTube algorithm is basically thee recommendation system that dictates which videos YouTube shows to its millions of users.

There is a common misconception amongst video marketers that YouTube only uses one algorithm to rank its content. That is completely untrue as YouTube functions on multiple algorithms that rank the content based on their different topics and where the videos surface (homepage, suggested videos, search results, etc.).

This recommendation system is specifically designed so that the viewers can get the videos they need. It brings videos to the viewers. I used to think that the system was implanted to promote certain videos and gain more views; however, that is not the case. When you open YouTube and land on the home page, the algorithm pulls up videos based on your engagement with videos on the platform. So for example, let’s say you recently watched a lot of videos on the stock market. When you open your home page next, many of the videos will be related to that topic.

YouTube Algorithm History

Over the years, the ranking system for YouTube’s content has changed quite the bit. Just take a look:

2005 – 2011: Clicks and Views

Jawed Karim, the founder of YouTube, says that the video platform was created in 2005 for one reason: to crowdsource the video of Janet Jackson and Justin Timberlake’s infamous Super Bowl performance. And so, till 2011 the videos on YouTube that received the most traction were the ones with the most clicks and views. The algorithm focused mostly on these types of videos. Soon, however, this proved to be ineffective because ranking solely on clicks and views lead to click-baits with misleading thumbnails and titles. More and more users came across such videos and were disappointed, so user engagement took a hit.

2012: Watch Time

Once the company realized that clicks and views was not the best criteria for ranking and recommending videos, they switched to basing it off of the watch time. When you come across a video that you like, I would assume that you would watch a majority of it because you enjoy the content. Once the creators realized that this was the new algorithm, they started making shorter videos so that their viewers would watch the whole video. Others, however, made their videos longer so people would spend more time watching it.

” YouTube didn’t endorse either of these tactics, and maintained the party line: make videos your audience wants to watch, and the algorithm will reward you.”

Paige Cooper, Hootsuite

And YouTube was right to not endorse these tactics because it ended up taking away the focus from the quality of the video. Creators were more concerned with the time people spent on their videos.

2015 – 2016: Satisfaction

To increase their user engagement they shifted their recommendation system to focus on viewers satisfaction. They used surveys and direct response metrics such as Likes/Dislikes, Shares, and Comments. Now, the algorithm was more personal and catered to each individual. It was no longer about rehoming videos just because many people had watched it. The focus shifted to finding the right video for that user.

“As a result, in 2018, YouTube’s Chief product officer mentioned on a panel that 70% of watch time on YouTube is spent watching videos the algorithm recommends,” says Cooper.

2016 – Present: Risky Content, Brand Safety, and Demonetization

As YouTube’s size and popularity has increased over the past few years, and this has lead to many issues as well. For starters, content moderation issues. Whatever the algorithm recommends has become quite the controversial topic. It has often concerned creators, advertisers, and even the government. On multiple occasions, platforms like YouTube have had to account for their algorithms at certain Senate hearings. Social media platforms impact our society greatly. So it is no wonder that the government has had to take steps like introducing the “Protecting Americans from Dangerous Algorithms Act.”

These controversies have lead to YouTube to claim that they will be more cautious with their guidelines. They stand to support diverse opinions, but also want to prevent the spread of misinformation that can be potentially harmful. Many creators have struggled to keep up with YouTube guidelines as they are not always transparent with their creators. Many creators fear that they may accidentally violate one of the company’s guidelines and suffer dire consequences such as strikes and demonetization. CEO Susan Wojcicki recently addressed this issue and released a statement saying that going forward their focus is to increase transparency on community guidelines with creators.

How Does the Algorithm Work Today?

YouTube’s algorithms today are dictated by two goals: choosing the right video for every viewer and convincing them to watch every video to the end.

As I mentioned earlier, there are three different types of algorithms:

  • Homepage
  • Search Results
  • Suggested Videos

So far, a majority of the traffic comes in from homepage and suggested videos. Search results, on the other hand, see more traffic for instructional videos.

What Determines the Algorithm?

So you may be wondering, how does YouTube know what video I want to watch and when? Well, there are three main things that ultimately determine how the algorithm works:

  • Personalization: Your history and your preferences
  • Performance: The video’s appeal, engagement, and satisfaction
  • External Factors: The audience/market for the video’s content

On your homepage, YouTube puts in a few videos based on things like click rate, likes, dislikes, views, comments, etc. If it engages and satisfies people of your niche, then it ends up on your homepage. It uses your past history to determine what types of videos you would be interested it. The suggested videos algorithm works a bit differently as it is mostly based on the types of videos you watch in one sitting. So if you watch a series of fantasy book reviews, then your suggested videos section will show more and more of that. As for the search results algorithm, that is based more on keywords and the performance of the videos.

While this was a lot to take in and go through, we have only scratched the surface. There is still so much to know and learn about algorithms. But for now you can rest easy knowing that YouTube is not secretly stalking you. They are simply using their algorithms to bring your the best videos.

Interested in more money content?