As part of a Tech Sales talent marketing campaign, I surveyed 180 associates in the tech sales team to identify what they value in a job, their online behaviors, and their reasons for choosing Red Hat.

This presented a problem. With so much feedback from the 180 interview participants, I needed a way to extract insights from the open-text responses quickly and accurately. This is where I first took a Python crash course, and sparked my interest in coding. I used the spaCy package and several online tutorials to derive top-level themes from the text using NLP, which we then drilled down into for further insight. This approach not only streamlined the analysis but also became a technique I continue to use in other projects today. It also inspired the below project as part of my diploma with Code Institute.

The Survey Sentiment Analyzer is a Python script that analyses survey data for sentiment and provides visual representations of the analysed data. This script allows users to connect to their Google Sheet data, fetch a specific category of data, analyse it for sentiment, and choose to either add the analysed data to their Google Sheet or visualise the data in a word cloud.

It’s designed to use basic natural language processing (NLP) techniques to analyze open-text data, particularly from open-ended questions in surveys. The application connects to Google Drive and Sheets via API, and is designed to help the user interact and understand survey responses at a high level, as well as export the data and wordcloud to use in stakeholder reports or anywhere else they may want to visually present the data.

I created this app to solve a problem I personally face often at work when analyzing such data. I’ve spent hours poring over open text survey data, trying to decipher common themes by reading alone. Equally, I can’t paste the data into an external text analysis service or word cloud maker, as it’s considered sensitive company data.

This app (if deployed on an employer’s IT-approved platform, or run locally on the user’s machine) helps users circumvent such data confidentiality concerns, and save hours on trying to read and understand bodies of open text data.

Currently, the app is linked to a spreadsheet containing publically available data: Female Empowerment Survey Data by Chandler Nunez on data.world. However, it can be edited to analyse any open-text survey data in a Google sheet based on user need.

Link to live app (temporarily offline)

Link to GitHub repo

Link to Google sheet