How Does the YouTube Algorithm Work in 2021?
YouTube was originally conceived as a social media meant for sharing videos. YouTube started gaining traction in 2009 and with viewership increasing exponentially, it meant the rise of the first generation of content creators. They became recognizable and some have become world-famous.
Advertising on videos appeared in late-2007. After the purchase by Google, YouTube left the traditional clickable ads behind. In 2009, YouTube established its trademark partner programs and established over 7 ways to advertise on the website.
YouTube’s content creators were the first step towards the influencer culture and advertisers, promotions and partner programs made it relatively easy for brands to market themselves and content creators to earn a living. Originally, a simple view count decided how much traction and popularity a video could get. A simple algorithm gave more ads on such videos and directed earnings to the uploader.
The YouTube scenario changed in 2016 with reports about terrorist organizations earning through YouTube’s revenue system and its most prominent creator Pewdiepie falling into a controversy regarding anti-Semitism. It began what is termed as the ‘Adpocalypse’.
Following the controversies, there was an industry-wide panic for the advertisers and companies who marketed through YouTube. Many pulled out of YouTube’s partner program and advertiser programs to keep their brand images intact. This led to YouTube overhauling their algorithms on monetization, search, suggestions, trending, and subscriptions. Ever since then, YouTube has never been the same. While YouTube successfully changed how its algorithms work, it took time thanks to controversies, rapid changes and huge blunders by YouTube with their new systems.
Thankfully, those days are behind us now and YouTube finally seems to be satisfying creators and corporations with how its algorithm works. There are simple tweaks to it every now and then but it is relatively stable in 2021. This article will focus on how YouTube’s algorithm works today.
The Current YouTube Algorithm
The rapid changes in YouTube’s algorithms over the past 3 years have left many confused. Even veteran YouTubers get confused about how it works and getting views has become harder in the present day scenario. That being said, it isn’t impossible to get millions of views on YouTube. YouTube, if done right, is still the best way to market a brand.
Why is it harder to get views on YouTube in 2021?
Keeping aside the changes that have been made to the algorithm, it comes down to a simple factor: YouTube’s mammoth-sized user base and creator base. More than 300 hours of videos are uploaded every minute to YouTube and hardly 5-10% of videos uploaded on YouTube reach a million views. That may seem like an insignificant figure but given the number of videos on the website, it’s a commendable number. The controversies also hurt the reputation and brand image of YouTube. How YouTube dealt with them was poor and the following panic hurt its creators. The mess took time to clean and even today, YouTube denies its mistakes. It made it harder to earn from the platform but it’s not exactly a dealbreaker. Once you’re set on YouTube, it becomes easier.
How does the Algorithm work?
The algorithm’s job on YouTube is simple. It allows users to discover videos. That’s all it does. Its job is to personalize the viewer experience on its platform. There are two ways for users to discover videos on YouTube:
- Search Functionality
- YouTube Suggestions
Search is the easier of the two. Search has hardly remained unchanged over the years and even through the controversies, YouTube did not mess up how searches work. Many YouTube creators and pundits feel searches are not important in the grand strategy. However, in an Internet filled with junk content, search is the dark horse of YouTube optimization.
Search is the easiest of the two to get right and is virtually dependent on keywords and watch time. The right title, thumbnail, and description can help you can rank higher on both YouTube and Google search rankings whereas the content drives up watch time which later affects suggestions best chrome extensions for youtube. Google gives preference to YouTube videos so make sure that your SEO is pinpoint if you want your videos to be discoverable.
The greatest advantage of optimizing your videos for search is the fact that it is virtually unaffected by changes to YouTube’s suggestion algorithm. There is no potential to ‘optimize’ search functionality further on the YouTube end and it is likely to remain unchanged in the coming years.
Advantages of Search functionality
- Safe in the long run
- Viewership increases with time as keywords mature
- Unaffected by controversies regarding privacy
- Precise results
Disadvantages of Search functionality
- Limited outreach
- Responsible for less than 30% of the views
YouTube suggestions have been evolving since the beginning. They are responsible for nearly 70% of the viewership on the website. However, YouTube suggestions are not limited to what it shows on the home tab. What most fail to understand is how vast YouTube’s suggestions are.
The home page is actually made up of various types of suggestions that YouTube gives to its users. It is a commendable webpage and YouTube deserves a pat on the back for its excellent plethora of suggestions available. Each of the suggestions seems the same but are different in their own regard.
The suggestions on YouTube can be divided into the following:
- YouTube Mixes
- <Channel> viewers also watch
- <Topic> tab
- Recently Uploaded
- “Up Next” Videos
However, this system of suggestions is not perfect. There is still a lot to be desired and irrelevant recommendations are not uncommon. However, that is a problem on the viewer’s end. Our focus is on leveraging these suggestion methods. The greatest downside to YouTube’s suggestions is a lack of control over viewer history. This is something no one will have control over and no one should, to be honest. YouTube had messed with this aspect and recommended only certain videos to viewers and faced unprecedented backlash from its own community and viewer base back in 2018. Hence, no matter how precise a video is, there is no guarantee that the viewer will be able to see your video. It just means you need to work harder for your views.
Coming to the suggestions, they tend to work through YouTube’s advanced and ever-evolving AI. All suggestions on YouTube are personalized for users by the AI meaning that no two users on YouTube have the same suggestions. Each type of YouTube suggestion uses the same AI in different ways. How AI works for each has been explained below:
Recommended: Recommended is what most people and businesses focus on. Getting your video to the recommended tab is a mammoth task. Mastering recommended indirectly helps you master the other types of YouTube suggestions. Recommended videos are selected based on most parameters that YouTube considers.The main factors involved in increasing your chances to end up in the recommended are keyword specification, watch time, thumbs up/down ratio, duration and series episodes. Remember that an ever continuing series always ends up on recommended. This fact is exploited by successful YouTubers and they always form playlists in their channels for videos. Recommended is usually for newer videos but YouTube’s algorithms recommend older videos based on a viewer’s search patterns.
YouTube Mixes: While often considered an unimportant factor, you should focus on youtube mixes. YouTube mixes help in the long run. Mixes are a playlist that YouTube AI randomly generates which consists of videos that the viewer has already seen with specks of unseen videos in the middle.While YouTube Mixes reroutes the old viewers of your video, their rewatching helps increase the watch time on a video substantially. Mixes are completely random and either contains the videos of a series or a channel.
<Channel> Viewers also Watch: This tab is considered a loss by many marketers, businesses and creators. It tends to drive viewers away from a particular channel towards a direct competitor. However, that is its USP. YouTube AI categorizes channels on the basis of their genres and the topics they cover. When YouTube senses that a user is interested in a particular keyword, it starts recommending them similar channels. This section suggests similar channels and topics. This section chooses videos just like the recommended section and is essentially just a rebadged recommended tab.
Yes, it takes viewers away viewers from your channel, but if your YouTube videos are AI-friendly, it brings viewers from other channels to yours. With good videos, you can literally steal viewers away from your competitors.
<Topic> Tab: Simple, straightforward recommendations based on a keyword. This section might be the same for different viewers since it recommends videos based on search history. This is a simplistic application of YouTube AI and bases its results on similar parameters as the recommended section. The topic tab can be exploited using the tags feature while uploading videos.
Recently Uploaded Tab: This is just another iteration of the subscriptions tab. However, it is different. Recently uploaded combines the subscription and recommended tabs and shows viewer video suggestions based on their recent watch history and the channels they’ve subscribed too. The catch, however, is that videos in this tab are less than a week old. Since most YouTube videos get 75% of their viewership within the first 7 days this tab is important. The recently uploaded tab is helped by factors like Upload frequency, channel strength, and watch time.
“Up Next” Videos: “Up Next” Menu is the menu shown in the right while watching a video. It was previously known as “Also Watch”. Up Next has been the source of YouTube’s headaches. Originally, it worked as YouTube’s recommended tab and recommended YouTube videos. Many creators feel that it should exclusively show videos of only the channel whose video is being watched. “Up Next” is still not perfected but follows the general rules of the recommended section.
For your marketing strategy, up next is potentially helped by a video series, video title, and relevance to the video being watched currently. It does give high preference to series based videos and will help linking viewers to the next video in the series.
Trending Tab: The trending tab is one of the oldest tabs on YouTube. Earlier marketing strategies relied on videos ending up on trending for their success. However, in the current scenario of YouTube and especially in 2021 trending is hardly relevant. It is a messy section and it is hard to get there. Apart from bragging rights, I have hardly ever seen this as helpful.
For a corporate channel, it’s even less important. Trending tab is more suited to the likes of individual creators and with the recommended section getting better trending is getting irrelevant with time. I believe with time, YouTube might phase out the trending tab. As a business, trending shouldn’t be your goal. You’re better off trying to hit the Recommended section.
Subscriptions Tab:The simplest tab on YouTube which shows a viewer about the uploads of the channel they’re subscribed too. However, what most businesses not do is utilize the story and activity features that the subscription tab offers. YouTube allows creators to post stories, and photos on the subscriptions tab. One can use these to hype upcoming videos, content, releases and updates. The subscription tab might not give you views but it can indirectly influence the brand image. Use it to your advantage.
Hence, the YouTube algorithm allows a video’s discovery through the search and suggestions functionalities. Using it to your advantage is up to the marketer and creator. Businesses usually have an advantage over content creators due to the availability of better resources and a pre-existing audience base which will further increase their presence on the social media giant. The algorithm does prefer certain factors for choosing a video to indirectly give it a boost. A well-made video is no longer good enough to increase viewership. A video’s success in 2021 is determined by its overall packaging, and while it seems biased against creators, in working, it actually provides a level playing field. It’s just that businesses have the advantage of having more resources than smaller creators.
What Should Your YouTube Channel Focus On?
Since the packaging of a YouTube video provides better traction, it becomes necessary to know what to pack it with. Given below are a few of the factors that make your YouTube videos algorithm-friendly.
Relevant Keywords: Keywords play a major role in YouTube search and topic suggestion sections. Find keywords before you post a video and making videos related to trending keywords can work wonders for your YouTube channel. This also includes the tags function that YouTube now offers.
Thumbs up/down ratio: The ratio between likes and dislikes can influence the AI. While not as important in the current scenario, YouTube’s AI does give preference to videos that have a higher thumbs up/down ratio.
Watch Time: The time period a video is watched. It does not mean the length of the video but rather how long do viewers watch the video. For example, a video might be 20 minutes long but if a viewer watches half of it, its watch time is 10 minutes.
Title/Description: The title and description of a video must contain relevant keywords if they’re going to rank in search results based on the topics involved. Longer descriptions can help Google rankings.
Series-based video production: The algorithm gives preference to continuous series. That is why popular channels, be it companies or individual creators, tend to have long-running series on YouTube with clearly labeled episode numbers. Series of videos never fail to end up in the recommended section.
Channel strength: A channel without subscribers cannot survive no matter how viral a video is. Subscriber strength is a sleeping giant. While all subscribers might not watch your videos it is a great way to generate a significant fan following. The AI gives preference to channels with high subscriber counts.
Off-platform promotion: Not directly related to YouTube but embedding videos on other social media can also help drive up its views which are also counted by YouTube. Example: 5-Minute Crafts major viewer base is on Facebook but they still have 60 million subscribers and billions of views, thanks to their relentless off-platform promotion.
Upload Frequency: YouTube algorithm still gives preference to upload frequency. It has to, thanks to the fact that millions of videos are posted on the platform each day.
All in all, this is how the present algorithm judges videos and recommends them to users. It is not perfect, but the algorithm is better than it was, say a year ago. The goal for creators and businesses on YouTube is to use this rough blueprint of the algorithm and optimize their videos accordingly to generate viewership and successfully market their channel. I must say though, that in an ever changing online landscape, the YouTube algorithm is as fragile as glass.
The above working is based on observation and experience. What goes on at YouTube’s headquarters is a mystery for us. The actors that influence AI have been covered carefully. One should remember that the algorithm works for the user and not for the creator. Hence, it is catered to viewer needs and it becomes essential for all channels to remember this rule while optimizing their videos. Even if everything is done right, a viewer might not appreciate your video and may push it to the ‘not interested’ list. Don’t let that discourage you. YouTube is a free service meant for people to share experiences. It is a social media and should be treated as one.
Good Luck on Your YouTube Journey!