Although YouTube is owned by Google, it’s a unique platform with a different approach to algorithms and discovery to Google Search.
YouTube’s systems take a viewer-first approach, evaluating a user’s preferences and interests based on the data it has about their viewing habits and pulling lists of videos unique to that user.
It also prioritizes signs of video quality, especially user engagement and user satisfaction, much more than metadata.
YouTube search is one of several systems that deliver videos, and while you can rank for specific keywords and search terms, it’s also weighted heavily toward personalization and user satisfaction.
Many videos may not perform well in search but will draw traffic from other discovery systems on the platform – and user journeys don’t include YouTube Search much of the time.
Ranking on YouTube is only one part of a complex relationship between you, your content, and your audience. If you want to build a successful YouTube strategy, then you must expand your thinking beyond ranking.
How YouTube Can Benefit SEO
YouTube has two key benefits for SEO.
YouTube is the second-biggest social media platform, with 2.5 billion monthly active users. An effective video strategy can drive brand awareness and engage your audience. It can also create a unique revenue stream through advertising and channel memberships.
YouTube videos can also appear in organic Google search results, and there is a tab for “videos” in Google Search. A video strategy can be critical to user experience, depending on the audiences you serve and how they prefer to interact.
I feel strongly that you should learn about YouTube’s algorithms even if you’re not committed to a YouTube strategy.
YouTube is a Google property, and even though there’s unlikely cross-pollination between YouTube’s and Google’s algorithms, YouTube is an excellent look at what Google prioritizes when it has access to every data metric and every user journey.
It’s a closed environment that it controls, which does not describe Search (despite its best efforts). Learning to succeed on YouTube can absolutely make you a better SEO.
How Do Videos Rank On YouTube?
To understand SEO on YouTube, you need to understand how the algorithm works and what its different systems prioritize. YouTube has multiple systems that deliver videos to users.
Each system acts using different sets of signals. There are general algorithmic priorities and then specific functions of systems that weigh different signals in different ways.
Generally, the different signals the YouTube uses fall under two categories:
Personalization.
Performance.
Under these categories, there are hundreds of individual signals.
Signals that fall under personalization include:
A user’s activity history tied to their account or browser.
A user’s activity history in the current session.
A user’s subscriptions, liked videos, notification preferences, and other interaction signals.
Search terms that users input.
Device type.
Time of day.
Signals that fall under performance include:
Total hours of watch time.
Average view duration/percentage of video viewed.
User satisfaction surveys.
Whether users ignore a video or click “not interested.”
Comments, likes, and subscriptions that a video generates.
User behavior while watching a video (skipping forward or back).
User behavior after watching a video (going back to their search, clicking on a new video, etc.).
For more information, watch the video below:
The Recommendation System
The first aspect of discovery we’re going to discuss involves video recommendations.
This term covers two distinct but closely related systems: a user’s homepage view and a user’s suggested videos view.
YouTube’s Creator Insider channel explains how the discovery systems work here.
The Homepage System
When a user visits YouTube, they have access to a homepage with video recommendations that are personalized for them.
This personalization is based on their activity history, what they like to watch, their subscribed channels, etc. The recommendations also consider signals of video quality and user satisfaction calculated by the algorithms.
To quote the video: “Home offers to deliver the most relevant, personalized recommendations to each viewer when they visit YouTube.”
The Suggested Videos System
Suggested videos (offered alongside the video a user is watching) are similar, grouped under the same umbrella of recommendations, but they use different signals. These recommendations priotize the experience of a user’s current session.
To quote the video: “Suggested offers viewers a selection of videos they’re most likely to watch next, based on their prior activity.”
While quality and user experience signals absolutely count in the suggested recommendations, they’re more heavily weighted toward satisfying a user’s immediate interest and intent.
This quote from the video is critical to understand, especially when we discuss different factors in YouTube’s algorithm later on: “YouTube’s recommendation system finds videos for viewers (rather than viewers for videos).”
The Search System
YouTube’s search system connects users with videos based on search terms they input, so it works a little more like Google Search.
It doesn’t weigh a user’s history quite as heavily to remain open to what they currently need.
Rather than trying to predict videos to suggest, it waits for user input and then uses available data (such as video topics, satisfaction metrics, and a user’s history) to serve results relevant to the query.
YouTube Search seems to be considered a separate feature. It prioritizes three core metrics:
Relevance: How well a video’s title, description, and content match the query.
Engagement: A video’s engagement statistics, such as watch time and other user signals.
Quality: YouTube also uses E-E-A-T signals, a familiar phrase for SEO professionals.
The Shorts System
The Shorts system is unique. While it uses many of the same principles as other systems, each user has a separate Shorts watch history from their long-form video watch history.
There may be some crossover, but for the most part, any personalization signals are separate between Shorts and other types of videos.
Shorts has its own tab, as well as sections on other areas of the site such as the home page, in Search, and recommendations.
For more information, watch Shorts and the Algorithm.
The Trending System
The Trending system is different from other systems in that it isn’t personalized. It shows the same videos to all users in a country.
Here’s what YouTube says about how Trending videos are chosen:
“Amongst the many great new videos on YouTube on any given day, Trending can only show a limited number. Trending aims to surface videos that:
Are appealing to a wide range of viewers.
Are not misleading, clickbaity or sensational.
Capture the breadth of what’s happening on YouTube and in the world.
Showcase a diversity of creators.
Ideally, are surprising or novel.
Trending aims to balance all of these considerations. To achieve this, Trending considers many signals, including (but not limited to):
View count.
How quickly the video is generating views (i.e., “temperature”).
Where views are coming from, including outside of YouTube.
The age of the video.
How the video performs compared to other recent uploads from the same channel.”
YouTube Algorithm Ranking Factors
The first thing you need to understand about YouTube’s algorithm is that it is “all about the audience.”
Recall this quote from earlier:
“YouTube’s recommendation system finds videos for viewers (rather than viewers for videos).”
It’s audience-first every time. Responding to a question about whether it makes sense to change the thumbnail or title of a video, a product manager at YouTube said this:
“When you change your title and thumbnail, you may notice that your video starts getting more or fewer views. And that’s generally because your video looks different to viewers, and that’s going to change up the way that people interact with it when it’s offered to them in recommendations. Our systems are responding to how viewers are reacting to your video differently, not the act of changing your title and thumbnail.”
1. Watch Time, Engagement, And Satisfaction
As explained in the video links above, the primary metrics that YouTube’s algorithm is optimized for are engagement and satisfaction.
Clicks and views are important in this consideration, but the most important is watch time modified by satisfaction.
Watch time can be represented in a couple of different ways:
The amount of time that a user watches a video.
The percentage of a video that a user watches.
Raw watch time is a pretty good indicator of whether or not users like a video. However, the percentage of a video they watch can be a better indicator.
In the analytics of a video, YouTube shows you the number of views, the watch time in total hours, and also a graph that looks something like this:
Screenshot from author, September 2024
This is a helpful visualization of the percentage of viewers engaged with the video and when they stop watching. You can use this graph to examine how successful your video is at grabbing and keeping attention.
This is critical information for you about the quality of your video in the eyes of your viewers, and the algorithm is using it as part of the formula determining how engaging and satisfying your videos are.
YouTube also trains its algorithms with satisfaction data, primarily gathered through surveys. The surveys aren’t a direct factor in an individual video, but the combination of watch time and satisfaction trains the algorithms to recognize high-quality content.
If you want to optimize for YouTube, or use video to enhance your SEO efforts, the first and most impactful rule is:
Make a good video that your audience wants to watch.
2. User Engagement
User engagement factors also matter for ranking.
How users behave during their watch time and after they are finished watching (combined with when they finish watching) goes into calculating video performance. These data sources include:
Whether a user skips forward during a video.
Whether a user moves back to rewatch a video section. (This can result in an automated “most replayed” section on a video’s timeline.)
How a user behaves after they finish watching. Do they:
Click on a new video in the suggested videos list?
Play the next video in the playlist?
Return to their search or homepage?
Close the session?
3. Video Title And Description
Just like websites in Google search, the title of a video is critical for helping users and search engines understand what content to expect.
However, it doesn’t act in the way you might expect. You should think about your title in terms of user experience first.
While it’s important to use keywords, your use of keywords will have much less impact than the CTR, watch time, and other engagement and satisfaction factors.
When it comes to your title, the most important consideration is how you set the stage for the video with your target audience.
Your title and video content are closely related, and you must ensure that the video delivers on the promise of the title – or at least begins to do so – quickly.
Optimizing your title for keywords can end up hurting you if it sets unrealistic expectations or makes the title less appealing.
Similarly, you should use descriptions in a way that helps users. The description should briefly describe the topic and what the user can expect.
It’s also a good place to put references from the video, links to other content the user may want to watch, and actions you may want them to take after watching, such as a link to your website.
A critical note for descriptions is that you should use them to add titled timestamps to your videos. Timestamps look like this:
0:00 Introduction
0:52 Chapter 1
02:02 Chapter 2
You wouldn’t want to use “chapter” – you would add a title that describes what that section of the video is about.
Doing this in the video description automatically adds sectioned chapters to the video timeline.
Google Search indexes these chapter titles and timestamps, so if you execute them carefully, Google can send users to a directly relevant section of your video. You can see this with the “key moments” feature in Google SERPs:
Screenshot from search for [how to train a dog], Google, September 2024As you can see in the next screenshot, the “key moments” appearing on Google Search are taken directly from the video description on YouTube:
Screenshot from YouTube, September 2024
This is powerful SEO. Do not underestimate this feature for “how to” SEO queries.
It’s a very compelling reason that video should be part of your organic SEO strategy because video results can show up in the “All” tab as well as the “Videos” tab in Google.
If you don’t add these yourself, Google may automatically create video chapters from the transcript. It’s best to control this process yourself.
4. Thumbnail
While they’re not a type of text metadata that algorithms can directly interpret, thumbnails are critical as part of the click-through and watch time formula.
Along with your title, they can make the difference between users clicking or not or finding a video satisfying or not, depending on the expectations set by the image.
I should also note that with Google leaning into multi-modal neural networks, it’s not beyond the realm of possibility that algorithms may soon understand thumbnails in relation to a video’s content.
Thumbnails are a ranking factor insofar as they’re a critical part of engagement. The algorithm will respond to how users respond to your video. So, your thumbnail needs to set your video up for viewers, both to attract clicks and set reasonable expectations.
In fact, thumbnails are one of the most important things on YouTube. They’re the first thing a user sees and likely the first piece of information they use to make a decision about whether to click, as they’ll likely parse the image more quickly than the title.
5. Video Content
Video content is, of course, the primary determiner of quality and what users respond to, so it’s what engagement and satisfaction data is based on.
However, there are some specific ways that a video’s content impacts ranking.
Watch Time: If you create a good video that viewers like, you’ll get positive engagement and satisfaction signals.
Captions and Transcripts: If you don’t provide captions, YouTube automatically generates them and builds a transcript for a video. It uses the information from captions to determine what your video is about, which makes your video script a helpful source of metadata.