A Detailed Guide to YouTube’s Algorithms
Understanding how YouTube’s algorithm works are crucial for your digital marketing strategy. You want to use the algorithm to your advantage, so you become more visible on people’s recommended videos and attract new followers. Optimizing your videos for YouTube combines riding on YouTube’s algorithm, producing high-quality content, and understanding what people want to watch.
We’re an SEO company from Los Angeles that knows a lot about marketing on YouTube, so we’ve written this article to help you understand more about it too. Here’s our crash course on YouTube’s algorithms!
Misconception about YouTube’s Algorithm
People think that YouTube is run by one algorithm, and in actuality, it’s run by many of them.
This article will focus on the two most essential algorithms—the homepage and suggested videos. Together, these two are what’s called the recommendation system.
The recommendation system should interest you because this is where most clicks and views come from. It’s hyper-personalized, attracting users with videos they want to see instead of popular videos. YouTube designed the homepage and suggested videos with user experience in mind, ensuring the user keeps clicking on the following video for as long as possible.
YouTube Ranking Factors: A History
What YouTube uses to rank videos has significantly shifted over time.
In 2011, the top-ranking factors were clicks and views. Quite obviously, this led to many clickbait titles and thumbnails that were often misleading, and YouTube had to figure out another way of ranking.
So, in 2012, the top-ranking factor was watch time. However, this, too, was flawed, as YouTube couldn’t measure the users’ enjoyment. It was also very easily abused as people could add interesting content towards the end and add some fluff at the beginning and middle parts.
So, since 2015, YouTube’s top-ranking signal has been viewer satisfaction. They want the users to feel good about what they’re watching, and any channel that doesn’t meet their standards gets ranked lower.
How YouTube Measures Viewer Satisfaction
But how exactly does one measure viewer satisfaction?
There are two main ways: surveys and signals.
You’ve probably seen one or two surveys before. It shows up on the top of your homepage and asks, “Is this a good recommendation for you? Help us improve YouTube”, with a video you might like and five stars.
On the other hand, signals are things found in people’s videos, like likes, dislikes, and notification bells. It can also be clicking the ‘not interested’ button.
How YouTube’s Homepage Works
Homepage videos are videos you want to watch, and they’re determined by both performance and personalization. Performance-based videos are videos other users with the same interests have enjoyed, so you’ll probably want to watch them too. On the other hand, personalized videos come from your watch history and will likely come from channels you’ve interacted with before.
As a marketer, you want to look appealing. If a new user sees you on their homepage and doesn’t know who you are, a well-crafted title and thumbnail can lure them in. Also, upload high-quality content consistently to have more shots on the homepage.
How YouTube’s Suggested Videos Work
Suggested videos show up on the right-hand side and are videos you want to watch after the current video. It’s affected by your watch history but also by the topic of the recent video. This helps YouTube recommend videos that you might not have seen before.
As a digital marketing agency in Los Angeles, we highly recommend keeping your current viewers on your channel. Make playlists (multiple videos on a broad topic) and series (multiple videos about a specific topic) to have that natural progression. Use end screens and make action calls to direct viewers to the following video. Use thumbnails that are readily identifiable as yours so people can find them in a second.
External Factors
How much traction your videos get depends on external factors, which are out of your control, and all you can do is work around them somehow.
The first is how niche your topic is. You can’t expect a million viewers to watch your content if only a hundred thousand are genuinely interested. Since there isn’t a lot of interest, to begin with, your content can get quickly buried by the algorithms.
The second is how broad your topic is. If you’ve got a thousand channels within the same subject, the user has to figure out how to share their time with each of the thousand channels or spend most of their time with a couple of their favorite tracks. You can figure out what usually happens. All you can do here is consistently upload all sorts of content, hoping you get noticed eventually.
The third and last external factor is seasonality, and not many people will watch Halloween-related videos in April. Additionally, the time people spend on YouTube significantly drops during the holiday season.