Visible App Store Reviews Analysis
Visible App Store Reviews Analysis
Role: UX Researcher / Data Analyst
Tools: R (tidytext, tidyverse, dplyr, janitor, tokenizers, wordcloud , stringr, lubridate)
Python (numpy, pandas, google play scraper) , GitHub
This Analysis is part of the Visible by Verizon Case Study. CLICK HERE to access the full case study.
Role: UX Researcher / Data Analyst
Tools: R (tidytext, tidyverse, dplyr, janitor, tokenizers, wordcloud , stringr, lubridate)
Python (numpy, pandas, google play scraper) , GitHub
This Analysis is part of the Visible by Verizon Case Study. CLICK HERE to access the full case study.
My Role
My Role
Scraped and filtered 10,000+ app reviews using Python and R.
Conducted sentiment analysis to uncover user pain points and positive experiences.
Built visualizations to show insights.
Translated findings into recommendations to improve onboarding.
Scraped and filtered 10,000+ app reviews using Python and R.
Conducted sentiment analysis to uncover user pain points and positive experiences.
Built visualizations to show insights.
Translated findings into recommendations to improve onboarding.
Problem Statement
Problem Statement
How do we capture the thoughts and opinions of hundreds of users without directly interviewing them?
Since I worked on this project independently, I turned to app store reviews as a source of user feedback. By analyzing these reviews, my goal was to discover both pain points and positive experiences within Visible’s onboarding process.
How do we capture the thoughts and opinions of hundreds of users without directly interviewing them?
Since I worked on this project independently, I turned to app store reviews as a source of user feedback. By analyzing these reviews, my goal was to discover both pain points and positive experiences within Visible’s onboarding process.
Some Parts Of The Data Prepartation
Some Parts Of The Data Prepartation
Getting Data into Python From Online
Getting Data into Python From Online
Version Coverage
Version Coverage
Filtering
Filtering
I pulled English‑language, US‑region reviews for Visible’s Android app.
I pulled English‑language, US‑region reviews for Visible’s Android app.
Quick quality control to see review counts by app version.
I conducted quality control to see review counts by app version.
I conducted quality control to see review counts by app version.
Filter reviews that only discuss aspects of the onboarding process.
Filtered reviews that only discussed aspects of the onboarding process.
Filtered reviews that only discussed aspects of the onboarding process.
visible_project = reviews_all('com.visiblemobile.flagship', sleep_milliseconds=0, lang = 'en', country = 'US')df['reviewCreatedVersion'].value_counts()onboarding_data = new_data %>% filter(str_detect(content, "signup|account|creation|onboarding|registration|setup"))Analysis
Analysis
After conducting Text Mining and Sentiment Analysis I visualized key patterns in the reviews, including:
After conducting Text Mining and Sentiment Analysis I visualized key patterns in the reviews, including:
Positive sentiment percentages for most common words
Positive sentiment percentages for most common words
Setup received the highest positive sentiment, with 60% of users expressing favorable opinions, highlighting it as a strength.
Login stood out as the most critical pain point, with only 25% positive sentiment, the lowest of all onboarding aspects
Setup received the highest positive sentiment, with 60% of users expressing favorable opinions, highlighting it as a strength.
Login stood out as the most critical pain point, with only 25% positive sentiment, the lowest of all onboarding aspects
Common positive words Included
Common positive words Included
Common negative words Included
Common negative words Included
Support
Support
Easy and Recommend
Easy and Recommend
Happy
Happy
Issues
Issues
Frustrating
Frustrating
Worst
Worst
% of Positive Opinions Line
Update
Service
App
Phone
Setup
Visible
Login
60%
25%
Sentiment Word Cloud
Sentiment Word Cloud
Account
Service
App
Phone
Visible
Setup
Common words associated with onboarding
Common words associated with onboarding
1
Account
2
Service
3
App
4
Phone
5
Visible
6
Setup
25%
60%
Results and Impact
Results and Impact
Setup received 60% positive sentiment, highlighting opportunities to reinforce this strength in the onboarding process.
Login appeared as the most critical pain point, with 75% negative sentiment, the lowest of all onboarding aspects.
Recommendations: Biometric login, Google/Apple sign-in, two-factor authentication (2FA), and a smoother checkout to reduce user frustration and increase trust.
Impact: Provided a research foundation for Visible’s onboarding improvements, with next steps of usability testing and user interviews to validate solutions.
Setup received 60% positive sentiment, highlighting opportunities to reinforce this strength in the onboarding process.
Login appeared as the most critical pain point, with 75% negative sentiment, the lowest of all onboarding aspects.
Recommendations: Biometric login, Google/Apple sign-in, two-factor authentication (2FA), and a smoother checkout to reduce user frustration and increase trust.
Impact: Provided a research foundation for Visible’s onboarding improvements, with next steps of usability testing and user interviews to validate solutions.
Limitations
Limitations
Google Play Skews only to Andriod users so IOS users may have different sentiment.
Google Play Skews only to Andriod users so IOS users may have different sentiment.
Reflection
Reflection