My Experience Doing Virtual Experience Program on Forage

7 min readMar 12, 2023

I recently participated in Accenture’s Data Analytics and Visualization Virtual Experience Program on Forage, a platform that offers virtual internships and other online learning opportunities. The program was designed to provide participants with hands-on experience in data analytics and visualization, and allowed me to gain practical skills and knowledge in these areas. In this article, I will share my experience doing the Virtual Experience Program on Forage, including what I learned, how the program was structured, and what I thought of the overall experience. But first, I will share to you what is Forage.

Forage

Forage is an open access platform designed to unlock exciting careers for students by connecting them with their company-endorsed Virtual Work Experience Programs.

Virtual Work Experience Programs are online programs built and endorsed by leading companies. These programs also open access for everyone, even international students. You can do them regardless of your visa or work status. They contain a series of resources and tasks designed to simulate the real-world experience of starting a career.

In a nutshell Virtual Work Experience Programs on Forage are designed first and foremost to help students. By completing a Forage Virtual Internship you will:

Better understand the diverse and exciting career pathways available to you, and,
Build the skills and confidence that will set you up for success as you move from the world of study to the world of work.

Virtual Internships are — and will always remain — completely free. You can also attempt any of our Virtual Internships with no prior work experience — try them out!

Project Understanding

Social Buzz is a fast growing tech unicorn that need to adapt quickly to their global scaling process efficiently. Accenture will running 3 months POC and the Data Team will focusing in an analysis of their content categories that highlights the top 5 categories with the largest aggregate popularity.

Social Buzz has over 100.000 posts per day, that can be assumed to have 36.500.000 posts per year. We will analyze the sample of this dataset from them.

Data Cleaning & Modeling

In this step, I will use Ms. Excel to cleaning and modeling to showcase my skills in Ms. Excel. Even though you can still use SQL to manipulate data, we can still use Ms. Excel to manipulate data to cleaning and modeling it. First we will look into the 3 tables that provided into us, “Content”, “Reactions” and “ReactionTypes” tables.

From this three, we can see that “ReactionTypes” and “Content” tables are dimension tables that explain the fact table, the “Reactions” table. Now we will filter all of it and check the value within it whether it already good or not.

First, we will look into “Content” table, We don’t need URL column because our analysis is specifically for content analysis, so delete it. And we look into Category column, there are several duplicates with quotation mark and there are no quotation mark. To make it same, we can replace it by doing shortcut key CTRL + F to find and replace it. We search “ and replace it with blank value so the category will be the same.

Replacing quotation marks with blank values

Now we also need to edit the column of “Type” into “Content Type” so we can know what kind of column it is. After this we delete User ID in “Content” and “Reactions” because they are irrelevant to our analysis. We also change “Type” in “Reactions” as “Reaction Type”. Then, we will look into null values in “Reaction Type” and Delete it.

Now we have cleaned datasets, with total rows of 24.574 rows (with header) of “Reactions” table. This table is our fact table to JOIN “Content” and “ReactionTypes” tables into it as our data modeling.

We will joining it with VLOOKUP formula. With VLOOKUP, we JOINing the three table like JOINs in SQL, we need to look into column with the same values, for “Content Type” and “Category”, we use “Content ID”, and for “Sentiment” and “Score”, we use “Reaction Type”. This is the formula in “Content Type”:

The Formula is =VLOOKUP(B2;’Content’!B:D;2;FALSE).

=VLOOKUP() is the formula,
[@[Content ID]] is the column for our values to look into the “Content” table,
’Content’!B:D;2 is the array or area we need to look,
2 is the number of column from left, I use 2 because Column1 is not used here,
FALSE is a command to look into it with the exact match.

Then we use VLOOKUP like the formula above according to their look up table, like “Content Type” and “Category” look into “Content” table and “Sentiment” and “Score” look into “ReactionTypes” table. We will have the cleaned data set like this:

Data Visualization & Storytelling

We can still use Excel to visualize it, but I will use Tableau Public so I can also share it here.

Before making a dashboard, there are several things that needs to be considered, what kind of information do we need to deliver and what kind of visualization we need to deliver it. I will list it here for what I need here:

Category Scoring, visualization of aggregate score value per category. I used bar chart here.
Top 5 Categories Popularity, visualize the share/percentage of the top 5 categories among themselves. I used pie chart here.
Posts per Month, how many posts there in a month. I use line chart to see the change over time.
Posts per Hour and per Day, how many posts is there per Hour and per Day. These 2 used bar chart too, because I wanted to see the number difference clearly.
Content Type of Popular Category, to help what type of content the user used the most, I used bar chart to see it.

I am also selecting Posts per Month in May, 2021, because it was time when the posts peaked. You can also look into it in Tableau Website here.

Present to the Client

Key Insights:

The top 5 popular content categories on Social Buzz are Animals, Science, Healthy Eating, Technology, and Food.
Healthy Eating is the top-ranking category with common theme of food, suggesting that users are interested in content related to health, wellness and food.
May 2021 had the most posts, with the most active days being Saturday, Sunday, and Monday.
The hours with the highest number of posts in May 2021 were 6, 7, and 8 in the morning, and 17, 23, and 0 in the evening/night.
The most popular content formats for each category varied, with Animals having more posts of photo and audio formats, Science having more posts of photo and video formats, and Healthy Eating having more posts of audio and video formats.

Impacts by Key Insights:

Consider creating more content related to Healthy Eating (the top-ranking category), Animals and Science (Factual and Real-Life) and tailor content to specific formats that are more likely to resonate with users.

By using the insights into the timing and format of user activity on the platform, Social Buzz can optimize the timing of content publication to reach the highest number of users and maximize user engagement. The insights into the most popular content formats for each category can also be used to tailor content to specific formats that are more likely to resonate with users, leading to increased engagement and overall platform growth.

Conclusion

In conclusion, attending Accenture’s Data Analytics and Visualization Virtual Experience Program on Forage provided me with valuable skills and hands-on experience in using Excel for data cleaning and modeling, and Tableau for data visualization and storytelling. Through this program, I gained insight into the workings of a large company like Accenture and was able to apply my newly acquired skills to complete a project. Overall, this program was a valuable learning opportunity and has equipped me with the skills needed to succeed in the field of data analytics and visualization.

Also, this is my Certificate of Completion.