This project aimed at helping the Pittsburgh-based non-profit organization, Bethlehem Haven, expand its digital influence by gaining valuable insights into their social media performance
In their quest to expand their digital influence, the Pittsburgh-based non-profit, Bethlehem Haven, has faced challenges in effectively connecting with new audiences through their social media channels. The ambiguous nature of what drives a successful social media post has proven to be an impediment. To overcome this hurdle, we developed a project aimed at deriving valuable insights through in-depth textual analysis of Bethlehem Haven's Instagram and LinkedIn posts. By recording specific variables such as likes, shares, and audience engagement for each post, we amassed a rich dataset for review. Furthermore, we incorporated additional parameters such as date of the post, word count, punctuation count, and content to discern emerging trends in how these factors influence post performance. To deepen our understanding, we incorporated various techniques of textual analysis, one of these being the application of NLTK's (Natural Language Toolkit) sentiment analyzer. Employing this tool enabled us to categorize posts into 'positive', 'neutral' or 'negative' based on their emotional undertone. We also conducted hypothesis testing to uncover potential links between these post features and key success indicators like likes or comments. Timing of posts, particularly in relation to important events on Bethlehem Haven's calendar such as their annual October gala, was another crucial factor we scrutinized in our analysis. Our overarching objective is to utilize these findings to help Bethlehem Haven enhance their online presence and ensure their social media content resonates strongly within their community. By fine-tuning their posting strategies based on these insights, they can ensure their mission and message reach a wider audience and foster more meaningful connections.
- Understand and analyze the performance of Bethlehem Haven's social media posts.
- Identify key factors influencing the success of social media posts.
- Provide actionable insights to enhance Bethlehem Haven's online presence.
- Optimize posting strategies for broader community engagement.
We collected data from Bethlehem Haven's Instagram and LinkedIn posts, recording variables such as likes, shares, date of the post, word count, punctuation count, and content details. Additionally, we implemented feature engineering using Part of Speech (POS) tagging to extract linguistic insights. This enriched dataset forms the foundation for our in-depth analysis. Data engineering techniques were also applied to ensure data quality and consistency.
We employed various textual analysis techniques, including Natural Language Toolkit's (NLTK) sentiment analyzer and Part of Speech (POS) tagging for feature engineering. These approaches helped us gain insights into both the emotional impact and linguistic characteristics of posts on the audience.
Hypothesis testing was conducted to explore potential links between post features (e.g., word count, sentiment) and key success indicators like likes and comments. This statistical analysis aimed to uncover patterns and correlations within the data.