ANLY482 Team wiki: 2015T2 TeamROLL Project Overview

HOME

ABOUT US

PROJECT OVERVIEW

PROJECT MANAGEMENT

DOCUMENTATION

Proposal

Midterm

Final

Overview

SGAG, one of Singapore's leading local humour content creators, maintains popular social media sites, an online website and a mobile application. Creating creative content on a daily basis, SGAG has garnered a significant number of followers on the various platforms. With an aim to achieve growth, SGAG hopes to leverage on their rich pool of data and derive valuable insights towards content creation.

However with limited resources, SGAG could not conduct a comprehensive analysis and harness on the big data available to them. This project aims to uncover valuable insights on SGAG’s content attributes in order to achieve audience growth. Using data gathered from SGAG’s facebook page for the year 2015, the team hopes to firstly, conduct exploratory data analysis so as to identify overall performance trends. Next, the team will be performing cluster analysis followed by sentiment analysis, topic analysis and content analysis. Lastly, the team will be building a regression model, which includes findings derived from the analysis conducted, in order to predict better performing future posts. With the insights gained, the team will be providing recommendations to enable data driven content creation, thus allowing SGAG to achieve their aim of greater growth.

About SGAG

SGAG is one of Singapore's leading local humour content creators. Distributing their creative content through multiple platforms, including popular social media sites, an online website, and a mobile app, the team at SGAG creates quality content daily to engage and entertain Singaporeans. With the goal to make "every Singaporean's day a better one", the company was founded in 2012 by two Singapore Management University undergraduates. As of today, SGAG has since gained a loyal following and have reached out to more than 300 000 Facebook users and 120 000 Twitter users, in addition to at least 200 000 users through other social media platforms, mobile apps and websites. Looking forward, SGAG aims to achieve greater growth and reach among their customers, especially for their target customers of Singaporean youths, working adults and young families between the ages of 18 to 34 years old.

Project Motivation

SGAG’s motto is “to make readers laugh at least 5 times a day, 365 days a year”. As such, SGAG places emphasis on the quality of their 5 daily posts, to ensure that readers will find their posts humorous with a local twist, a good-natured piece of fun without any intention to hurt. Much of their content focuses on a funny stereotype of everyday Singaporean life which locals are able to identify with. Although SGAG has been very successful thus far, it also recognises that the online content space is very competitive, with newer players such as "SMRT Feedback" joining in the fray to generate local humour content. As such, SGAG needs to evaluate and improve their content strategy to ensure they stay entertaining and engaging to their audience. However, SGAG faces a few challenges in understanding and thus, leveraging on, their past success factors, which may be summarised as follows:

1. What are the characteristics of a “great” post? SGAG has so far thrived on an intuitive understanding of their customer's content preferences. However, SGAG does not have a concrete or clear picture of the kinds of attributes which they can work on to make a specific post a "great" one.
2. What is audience sentiment on "viral" posts? Are they reacting in a positive or negative manner? SGAG is concerned that "viral" posts become popular because they receive a lot of "hate", which goes against their content philosophy which is to make people "laugh", a positive emotion. Currently, they do not have easy visibility on this aspect.

SGAG hopes this project will be able to utilise a rich pool of historical data to derive insights into the concerns posed above, so that SGAG would be better able to formulate a more relevant content creation strategy.

Project Objective

The final goal of this project is to offer useful insights for SGAG to formulate a better content creation strategy moving forward. To measure the effectiveness of their content strategy, and at a more granular level, the effectiveness of each individual post, SGAG operationalises effectiveness as "growth" which is defined by an increase in 1) Number of fans, 2) Audience reach, and 3) Engagement with audience members. This last indicator is further measured by the number of times audience members perform actions such as “likes”, “comments”, “shares”, “retweets” or clicking on links to find out more about the content SGAG has to offer. To do so, we attempt to answer the two main challenges posed by SGAG in a concrete, data-driven manner by performing an in-depth analysis on SGAG's historical data. More specifically, we attempt to address the following analysis requirements:

1. To be able to understand whether a post is popular in a “positive” or “negative” manner
2. To assess the role of content layout and design in improving popularity of posts.
3. To develop a list of common topics and be able to understand the role of topic-selection in affecting the popularity of posts

Data collection and description

Our two main datasets are: Facebook Insights Data Export - SGAG - Page Level, and Facebook Insights Data Export - SGAG - Post Level. The datasets are sponsored by SGAG and extracted from the Facebook Insights tool. A year's worth of data from 2015 was extracted. Although SGAG also obtained similar data for the same time period from Twitter through Twitter Analytics, this would not be the focus of our project for the present time.

Facebook Insights Data Export - SGAG - Page Level

This dataset captures key performance indicators of SGAG at the page level. These include variables such as lifetime total likes, new likes, unlikes, number of engaged users, reach, organic reach, number of clicks on content, and number of negative feedback, on the daily level, or aggregated to form weekly and 28 days measures. This dataset also captures information regarding the demographics of SGAG's customers, their ages and gender, as well as their location in terms of countries and cities.

Facebook Insights Data Export - SGAG - Post Level

This dataset similarly captures key metrics of SGAG, but at the post level. Many variables found in the earlier dataset are also reflected in this dataset, but at the post level. We propose that this dataset be our main point of analysis for this project, with the earlier dataset utilised as a supporting analysis.

Work Scope

Our proposed work scope will focus on the main content distribution channel SGAG currently uses, which is Facebook. This would be where SGAG garners the most reach and engagement from their target audience. We will also be conducting our analysis based on historical Facebook data for the year 2015, which is suitable due to it being relatively recent. A step-by-step breakdown of our proposed scope of analysis is as follows:

1. Data Collection – Collect Facebook data for the year 2015 to be analysed, from SGAG
2. Data Preparation – Clean and transform data into a readable CSV for upload
3. Exploratory Data Analysis - Identify overall performance trends
4. Cluster Analysis – Perform segmentation of Facebook posts based on their performance in terms of total reach and engagement level (likes, shares, comments)
5. Sentiment Analysis – Identify differing sentiments based on posts and clusters
6. Topic Analysis - Generate and identify topics based on posts and clusters
7. Content Analysis - Identify key design attributes based on posts and clusters
8. Regression Modelling – Build a regression model that includes success factors derived from analysis, to aid in predicting better performing future posts