In the above screen shot, the search query is SEO Tips, and the video that occupies the top spot is "Google On Page SEO Tips and Strategies."
This video can get a high click through rate, which results in high video exposure. This exposure will increase the number of video views and possibly the earnings (if you have ads in those videos or a referral link to your business website for customers to purchase products).
This article attempts to examine different factors that affect the video ranking in YouTube search and recommends the best course of action that could significantly help the marketing of the video in YouTube.
Formulation of Possible Factors to be Examined
Before we gather actual measurable data in YouTube to confirm the significant factors, it is important to list the possible factors that can affect the ranking of a YouTube video. These can be separated into three categories.
First, there is the video’s relevance. This is a measurement of how relevant the video is to the search query. In YouTube, since the content is video, the best measurement of relevance is the video title.
A good way to measure the relevance is to count the number of words in the actual video title that match the actual query. For example, say the search query is "Online tutorial on Coffee brewing." Some example video titles can be rated as follows:
1. Coffee Brewing Online = 3 words match the query. So it is 3/5 = 60% relevant. The query contains five words.
2. How to create delicious Italian Coffee = Only one word matches the query. So it is 1/5= 20% relevant
3. Coffee Brewing – an Online Tutorial = 4 words match the search query, which means it is 4/5 or 80% relevant.
The second factor that may affect a video’s search position is User Engagement. This is the measurement of users’ engagement and their interaction and perception of the video.
Measurable factors include the following:
1. Favorites - the number of times the video has been added as "favorite" by the YouTube viewers.
2. Number of comments - the total number of user comments on the video.
3. Number of likes - the number of viewers that "like" the video. This replaces the old YouTube five-star rating system.
The third factor is Trust and Authority of the Video Owner. Similar to the way websites are ranked in Google, the trust and authority of the video owner might also affect the video ranking in YouTube. Of course, YouTube, in the effort to avoid spam, will most likely check the background of the YouTube video owner, and could use that information to assess the trustworthiness of the video.
Measurable factors under this category include the following:
1. Membership age – the number of years that the video publisher has been active in YouTube. The older the membership, the more trusted is the video owner. This concept is similar to the age of domain in Google search engine ranking factors.
2. Total Channel Views – the total number of viewers that visit the video publisher/owner channel.
3. Total number of subscribers - this is important; the higher this figure, the more authoritative is the video owner.
For information purposes, the data for the factors below will be gathered from the following sources:
1. Keywords in title – the <title></title> element of the video URL in YouTube, excluding the word "YouTube."
2. Membership age – the formula to compute this is as follows:
Age = 2010 -Date that the owner joined YouTube
3. Number of Subscribers and Total Channel Views – these data are available from the YouTube Channel of the video owner.
4. Number of Favorites, Number of Comments, and Number of Likes – can found in the actual video URL. Click the drop down beside "Views" (red box below):
Five sample keywords will be used in this analysis, which are as follows:
1. Tips to play piano
2. Guitar lessons
3. Laptop repair
4. PHP developer
5. Mountain climbing
Raw data will be tabulated in a spreadsheet. You can download the raw data for yourself.
All data gathered are numerical in nature and converted to a 0% to 100% rating scale by taking its percentage based on a total.
The orange columns are the final data to be analyzed.
Since all data are of the objective type (numerical), the percentage importance of each factor can be analyzed using MS Excel Solver. Because of this, a summary table has been created using the raw data you downloaded previously. The %importance column (in pink cells) is blank because it is still unknown (see the big screen shot).
Using Solver with the following constraints:
$E$10 = 1
All variables in cell $E$3 to $E$9 should have values defined by:
<= 1 and >=0
Here is a sample screen shot of the Excel solver with target and constraints:
Since there are five keywords, the correlation average is taken from those five samples, and the result of that average is placed in cell $N$2. The goal (target) of the solver is to set this average to 1; this will correspond to decreased ratings from Position 1 to Position 240 in YouTube search. In layman’s terms, it simply asked: "How much would be the % importance of each measurable factor such that the decreasing trend rating of scores is nearly perfect from top to bottom?"
The Solver spreadsheet with the optimal value of "%importance" can be downloaded here.
At 76% accuracy (other optimization engines might find a better solution or increased accuracy), the following are the surprising results:
First, note that keywords in title tag is the most important YouTube ranking element in their search. According to the analysis results, it has around 82% importance.
This is crucial, because the first thing that the YouTube user will see in the search results is the title tag of the video. Keywords are also bolded in the title tag, indicating their importance.
The next important factor is "number of years as a YouTube member," at 16%. YouTube just started in 2005, but with the exponential growth of video owners and some just-created channels for spamming, one of the best factors to use for checking user trustworthiness is the length of membership.
Old members tend to be authoritative because of their experience and reputation as YouTube video owners.
The third most important video ranking element in YouTube search is the number of favorites for that specific video. A higher figure here indicates that the quality of the video content is high.
This is also a more reliable measure of quality than "likes" or the "star" system because YouTube users are actually "adding" the videos to their own channel if they tagged them as a "favorite."
The last factor with an importance value is the number of channel subscribers for that video owner. Along with age of membership, a high number of subscribers is a good indication of being an "authority" and a trusted source of information.
Surprisingly, "number of comments," "number of likes" and "total channel views" did not turn out to be factors in the YouTube video search ranking. Why? Perhaps YouTube does not trust user engagement factors too much.
User engagement only comprises an estimated 2% importance in YouTube video ranking system. It is obvious that:
- YouTube might not be trusting factors that can be manipulated easily (for example the "likes," "number of comments" and "total channel views." This will prevent spammers from promoting their video to the top of YouTube search results using artificial methods.
- YouTube places more trust in the "authority" of the video owner by determining the number of subscribers as well as the number of years as a YouTube member.
Recommendations to achieve the best video ranking performance in YouTube
Based on the data analysis results, it is recommended that you do the following:
1. Create original and high quality video content. This will increase the number of users that will add your videos as their "favorites." This will also increase the number of subscribers to your videos. Both of these factors can have a significant effect on how your video is being ranked in the internal YouTube search.
2. Stay longer and consistently submit videos to YouTube. This will attract more subscribers in the long run, and while you "age" in YouTube, it will earn you some reputation and authority that will help your videos to rank better in YouTube search result.
The real/exact algorithm used by YouTube is not publicly known or published. The analysis results and recommendations discussed in this article will be a good start to help you effectively promote/market video in YouTube.