Last November, a YouTube channel owner in Mexico contacted me to design and implement a low-cost Cloud-Based Analytics Solution (e.g., a Recommender System that complements the channel's analytics already available) that helps him to select the best content to be uploaded, namely, videos, descriptions, stories, etc.
This customer is passionate about Bolero music in particular. He started the channel as a hobby a few years ago. Nowadays, however, he's thinking of becoming a more professional YouTuber and helping to spread this lovely music genre globally and, in the process, adding as many subscribers as possible to his channel. For me, tackling the challenge of joining the old (above) with the current cutting-edge technology (below) was a fascinating experience.
After several Google Meets and much discussion, we distilled the following four solution's main features:
A solution easy-to-explore that would offer, with a few clicks, a meaningful and insightful experience to the user
Input data AUTOMATIC-UPDATE
TIME EVOLUTION display of relevant metrics corresponding to the last twenty-eight days (minimum)
FILTERING CAPABILITIES; the user would be able to download, as CSV, Excel, or Google Sheets, relevant summaries, etc.
Before moving forward into the solution design and its implementation, I due to answer my EXCEL-MINDED client three basic but fundamental questions: what is Data Analytics? How many Analytics types are there? Why are Cloud-Based Analytics Solutions relevant?
Data Analytics comprises examining data sets (no matter their size and complexity) to find trends, distill knowledge, and draw conclusions about the information they contain.
There are three types of Data Analytics:
Descriptive Analytics, answers the question, “What happened?”
Predictive Analytics is used to predict future trends or events and answers the question, “What might happen in the future?”
Prescriptive Analytics takes into account all possible factors in a scenario and suggests actionable takeaways. It answers the question, “What should we do next?”
A Cloud-Based Analytics Solution involves the deployment of SCALABLE cloud computing (storage, connection, preparation, AUTOMATION, etc.) with powerful analytic software to distill knowledge and extract new insights from data and information available. These technologies and techniques are widely used in several industries to enable organizations (in almost real-time) to make MORE-INFORMED business decisions.
With the previous established, the next step was, taking into account budget limitations, to pick up the services and tools with which to build the required solution successfully.
Again, the answer was to use some of the Google Workspace and Google Cloud Platform services and tools:
Google Drive to store and organize part of the data
Google Looker Studio (Data Studio) and one of its connectors (YouTube Analytics CONNECTOR)
Google Sheets and Google Slides, etc.
Most of these applications/tools are free of cost, so the limited budget was no more an issue.
The BI software Google Looker Studio was included to serve the project's results as compelling dashboards containing fully interactive visualizations. Watch the video below for a clear description of this powerful tool, part of the Google Cloud Platform.
Another solution's crucial element included was the YouTube Analytics connector available to Looker Studio. It has the following key features:
Direct access to the most relevant features (total watching time, number of visits, audience data, etc.)
Access data from the last 28 days directly from the channel
Updated data is available every 12 hours
No data preparation was carried out to achieve the specific objectives of this project. However, if required, it is possible to plug in the connector with BigQuery (another powerful service of the Google Cloud Platform) and connect the resulting data table directly with TRIFACTA and prepare the data; the refined dataset can be published back in BigQuery and be accessed directly from Looker Studio. This type of configuration will be discussed in detail in future posts. The resulting Cloud-Based Solution architecture, at a high level, is shown in the image below.
.
Once the connection between Looker Studio and the YouTube channel has been established, a list of the available features or variables is displayed, which can be used immediately to construct maps, tables, time series (with trendlines) and bar plots, etc., implement filters with which slice and dice the data and explore it, by country, title, subscription (true/false), etc., and extract with a few clicks actionable insight from all the data available, pursuing the best user experience.
Also, it is possible to access additional data from Google Drive (or Google Sheets), embed and showcases a presentation from Google Slides, and download relevant content into a Google Sheet or other formats like CSV and EXCEL, if needed.
The figure above depicts one of the solution's dashboards. All have been planned and designed to offer the best possible experience to the user: easy-to-navigate and easy-to-digest compelling and insightful visualizations, allowing the user to get the most value out of the data available with just a few clicks.
Summary
Wrapping up:
A fully interactive and EASY-to-DIGEST compelling visualizations Cloud-Based Analytics Solution was successfully implemented that assist the channel owner in selecting the best content to be uploaded by country, title, etc.
The YouTube connector allows input data AUTOMATED-UPDATE every 12 hours.
FILTERING CAPABILITIES that allow slice-and-dice of the data and then download the results were also implemented.
A Time-Series interactive chart (with trendlines) is available to display and analyze the metric Total Watch Time (TWT). This feature allows the user to anticipate trends and apply corrective actions proactively.
Are you interested in a One-on-One USER TEST of this solution?
In future posts, I’ll continue discussing more real-life relevant use cases. Please, don’t miss them out. Leave your comments below, and kindly share and contact me if you require additional information.
コメント