Accenture Virtual Experience on Forage
- Abdul-Sherrif Saaka
- Sep 22, 2022
- 2 min read
Updated: Oct 24, 2022
Welcome back! This new project is special to me because it was a real testament to how far I have come. I tried this exact virtual experience about 3 months ago and I couldn't even identify the business problem. Fast forward to now and I handled it pretty well.
I applied Excel and Power Query to clean and analyze the data, Power Bi to visualize and Google Slides to present the entire process.
Background
In this project I take on the role of data analyst in a Proof of Concept for Accenture to help a social media client adapt to massive growth. My specific task is to provide insights regarding the top 5 most popular content categories on the client's platform.
Data Processing
To begin this project my team was provided with 7 datasets all in csv format
1.1 files

Now to find and create the dataset I needed to answer the business question I loaded all the datasets into power query. From close inspection I see that the Content table contains some of the data I need for my model. I go through steps to clean the data from wrong, empty and duplicate values in the primary key field . Popularity of content was based on reaction score from 0 to 75. The higher the number the more popular a post is and the more popular the category of the post.
1.2 Reaction types key table

To get all the reactions with corresponding scores in one table I do a left outer merge between the Reactions table and ReactionTypes table. After that I go through my cleaning processes again to make sure data is ready for modelling and analysis. All other tables are not needed to answer this particular question so I remove them. Now I have just two tables; Content and Reactions
Data Modelling
The model for the two tables with Power Bi is simple and straightforward. content_id is the primary key of the Content table and I connect that to content_id on the new Reactions table where it is a foreign key
2.1 Data Model

Visualization, Insights and Presentation
After modelling the datasets and analyzing it is time to answer the client's question with accompanying visualizations in a PowerPoint presentation below
3.1 Presentation to clients
Conclusion
This task was not exactly difficult but for me it felt so good to realize that I have built real data skills. That I could look at data and see a story and that I could tell that story. It was a proud moment for me. But job's not done. We keep learning and in the spirit of learning I will always be open to suggestions. Thanks for reading!
.png)




Comments