Portfolio
I’m a firm believer that great things happen when curiosity meets hard work. Here, I’ve brought together projects that reflect my journey of learning, growing, and solving real-world challenges with data.
/ Portfolio
/ Youtube Analysis for Channel Optimization
Youtube Analysis for Channel Optimization
Github's repository:
Final Report:
YouTube was the second most popular social network in 2023 (globalmediainsight), boasting 2.7 billion monthly active users.
​
Brands and entrepreneurs have gravitated towards this platform for two primary reasons:
I. to increase their visibility by growing their subscriber count and views, and
II. to earn revenue through channel monetization.
​
By using the STAR method, we will analyze data and strategies for YouTube creators and businesses to achieve these key objectives.
​
We programmarly downloaded the data from Kaggle using the Kaggle API.
Link to the dataset: click here
​
Then, we start by understanding and preprocessing the dataset. While we were cleaning the data, we found out that the "channel_type" column does not match the values specified in the dataset description, which is supposed to represent the type of YouTube channel (e.g., individual, brand). Therefore, we will drop that column to streamline our future analysis. ​
​
RESEARCH QUESTIONS:
-
What are the most popular content topics on YouTube in 2023?
-
How does the number of uploaded videos impact channel growth?
-
How does content diversity affect channel performance?
-
What are the demographics of engaged YouTube audiences?
-
Is there a correlation between the number of subscribers, views and remuneration from YouTube?
​
PROPOSED HYPOTHESIS:
-
YouTube covers many topics, but due to its popularity in the social network sphere, the top three most popular topics on YouTube are likely Entertainment, News, and Education.
-
The more videos you upload on the platform, the more YouTube's algorithm considers you as an active member and suggests your channel more frequently to a larger audience.
-
Diversity is useful for covering a range of topics, but on YouTube, it's often better to focus on one niche and make a significant impact in that domain.
-
The United States has the largest audience on YouTube. Therefore, content creators and brands can adjust their content creation strategies to align with the United States' YouTube content models, including trends and using English as the primary language, to access a larger viewer base.
-
Yes, the more subscribers and views you have, the higher the remuneration from the channel.
​
DATA ANALYSIS
We decided to go with the metrics that interested us the most: views, subscribers, earnings, category, uploads, country.
​
Starting by checking the distribution of Youtubers by categories.
From this pie chart, we could observe that the top 3 categories on Youtube are topics about Entertainment, Music and People & Blogs for the year 2023.
​
Most content creators and brands dive in these popular categories due to the success of these contents in terms of the number of subscribers, as shown in the following graph.​
​
Next, does uploading more content videos lead to better channel performance, resulting in gaining new subscribers and increasing exposure through higher view counts?​
​
The bar chart indicates that ABP NEWS got the highest total number of videos uploaded on Youtube untill the year 2023, followed by GMA Integrated News, and TV9 Bharatvarsh on the third position. ​
​
But will they keep their actual rankings in terms of number of subscribers and views? We'll find out shortly. ​
The distribution of the top 10 YouTubers in terms of views and subscribers differs significantly from the classification of the top 10 YouTube channels based on uploaded videos. For instance, T-Series performs well despite not ranking at the top in terms of uploaded videos.​
​
However, is there truly no correlation between the number of videos and the number of subscribers? Let's analyze it.
The two scatterplot charts reveal insights as we observe negative relationships between the number of uploads and the number of views and subscribers. This leads to the conclusion that quality triumphs over quantity on YouTube. Having a high number of uploaded videos doesn't necessarily attract more audience views and subscribers.
​
However, we can observe outliers in the graph, indicating that some channels were created early and could have gained much more popularity.​
​
Additionally, there are clustering points in these visualizations that require deeper analysis to determine potential patterns or subpopulations for future project.​
​
Next, we also wanted to see if channel diversity affects channel performance.​
By examining this plot, which is ordered by median, we notice that the center of distribution is quite similar across most categories. However, "Trailers" stands out with a higher median than the others.​
​
Furthermore, we observe variations in the Interquartile Range and the number of outliers among these categories.​
One factor that may contribute to these variations is the sample size in this specific dataset, which already includes well-performing channels. It's possible that categories with less variation also have fewer records.
​
Indeed, the top-performing channels are those with fewer records in the dataset. This aligns with the concept of a few "giant" channels dominating the niche rather than numerous big channels performing well.
​
Certain categories have a high number of records, such as "Music," "People & Blogs," "Entertainment," and "Gaming." This suggests a different public behavior pattern, with numerous highly-performing diverse channels rather than a few massive ones dominating the niche. It's notable that these categories also tend to have more outliers and a larger Interquartile Range (IQR).​
​
Niche-focused channels tend to experience faster subscriber growth due to targeted content that resonates with specific audiences. This targeted approach fosters long-term sustainability by establishing expertise and authority in a niche, leading to loyal and engaged subscribers.​
​
In contrast, channels with diverse content may attract a broader audience initially but may struggle to maintain audience engagement and retention over time.​
​
Thus, while diversity has its merits, niche-focused channels often excel in subscriber growth and long-term success on YouTube.
​
Next, we would like to check the demographics of Youtube channels and audience, in order to optimize future marketing strategy with the corresponding audience.​
​
When analyzing this dataset, we discovered an interesting presence of missing data for specific countries. The absence of this data collection may significantly impact the number of channels in the top countries.
​
With the exception of the African continent, all regions seem to have a reasonable representation, albeit with limited entries. The US and India dominate in hosting the majority of channels, while other regions may not even reach a dozen.​
​
Both countries boast large populations and robust media industries. With Hollywood in the United States and Bollywood in India being the world's top film capitals, many creators have turned to YouTube to share entertainment content.
​
Market segmentation is crucial for developing effective business strategies, and the same principle applies to activities on YouTube. Understanding your target audience is essential. Therefore, let's analyze the average number of subscribers across different countries.​
​
The comparison of the average number of subscribers between countries reveals no significant differentiation among them in the realm of high-achieving YouTube channels.
​
However, we can gain a more comprehensive view of the situation by examining the main category breakdown by country.
Targeting a broad audience by focusing on popular categories like Entertainment, Music, and People & Blogs is advisable, given their widespread popularity across highly populated countries.
​
However, considering regional preferences is crucial, with Music channels being favored in America, Entertainment channels in Asia, and Europe showing diverse interests.
​
Content creators should also capitalize on the concentration of channel hosting in America and India, tailoring content to suit these audiences while exploring niche categories in Europe.​
​
Ultimately, a balanced approach that combines popular global categories with regional nuances can help maximize reach and engagement on a YouTube channel.
​
Finally, we aim to analyze correlations between the number of subscribers and views and their impact on remuneration from YouTube.​
​
While there is a positive correlation, it's not exceptionally strong, suggesting that a high number of subscribers and views does not guarantee proportionately high income alone.
​
Other factors like video duration, user interaction, and ad placement can also impact channel revenue.
Successful YouTube channels often have various revenue sources such as merchandise sales, sponsored content, affiliate marketing, and fan contributions, reducing their dependence on traditional advertising revenue or income from views and channel subscriptions.
​
CONCLUSION
The project ends here. You can view the Tableau visualization at the following link: Tableau
To check my other projects: Portfolio