Social Buzz Data Analytics and Visualisation Project

Gerhardt - Jan 18 - - Dev Community

INTRODUCTION

In this project, I carried out an extensive analysis for Social Buzz, a social media and content creation company that has scaled quicker than anticipated and needs the help of an advisory firm to oversee its scaling process effectively.

Due to the huge amount of data they create, collect and analyse they are now willing to bring in an external expertise to help with the management.

DATA PREPARATION AND CLEANING
Upon reading the brief from the social buzz, I understood the requirements needed to be delivered for this project. These requirements are an audit of big data practice, recommendations for IPO, and analysis of popular content.

I was provided with 7 datasets and a data model.
So, firstly I used the data model to identify which datasets will be required to answer the business question - which is to to figure out the top 5 categories with the largest popularity.
Below is the image of the data model:

Image description

Next, I needed to ensure that the data was clean and ready for analysis…
To clean the data, I removed rows that were irrelevant to the analysis and had values which are missing, changed the data type of some values within a column, and removed columns that are not relevant to this task. After, I joined the relevant columns from your Content data set, and then the Reaction Types data set using the VLookUp formula.
Here are images of the datasets in Microsoft Excel:

Image description
Image description
Image description

ANALYSIS

The Analysis was done to get some insights and answers to some imperative questions and to get an understanding of the dataset.
This was done by the use of catchy visualizations.
Before that, I’d like to reiterate the problem and what the analysis should find.

PROBLEM
Over 100000 posts per day
36,500,000 pieces of content per year

So the analysis should help us find social buzz’s to 5 most popular categories of content.

Below are images of the insights, top five content categories and top five aggregate by popular score.

Image description

Image description

Image description

ANALYSIS - Studying and healthy eating are the two most common categories in the aggregate score.
This shows that people prioritize
"cognitive" and "longevity" content the most.

INSIGHTS - Food is the most common theme of the top 5 with
"studying" ranking the highest. You could use this insight to create a campaign and work with cognitive and mental development brands to boost user engagement.

NEXT STEPS - This ad-hoc analysis is insightful, but it's time to take this analysis into a larger-scale production for a real-time understanding of your business.

This is the conclusion and I appreciate you taking out time to read it.

. . .