Into the Glossier Reviews : Insights from Top 10 Products

Posted on Feb 2, 2020
The skills I demoed here can be learned through taking Data Science with Machine Learning bootcamp with NYC Data Science Academy.


Glossier is a direct-to-consumer millennial makeup brand that has grown rapidly in the past 5 years. Founded by beauty blogger Emily Weiss, the Glossier brand was launched from her original beauty blog, Into the Gloss. Glossier focuses on helping consumer build an easy, simple beauty routine. Emphasizing their mission to "democratize beauty", the brand leverages community and active consumer feedback.

Given the brand's focus on understanding its consumer base, I wanted to explore the reviews of Glossier's most popular products to see if I could better understand consumer engagement and feedback. A link to my project's code can be found in my github repository

Data Scraping

Arbitrarily defining "popular" as products that have more than one thousand reviews, there were 10 popular products available as of January 22nd, 2020. 

There are several components of a review for Glossier products, split into two main parts: 

  1. Customer Description 
    • User name
    • Location
    • Skin type 
    • Skin shade
    • Age range
  2. Feedback 
    • Date of review
    • Number of stars
    • Review title
    • Review text
    • Bottom Line - would you recommend this to a friend?

I utilized Selenium, a Python web scraping tool, to scrape the reviews of each popular product. This resulted in 18,269 reviews scraped.

Data Preprocessing

After scraping all popular products, I had the reviews from the following product names: 

  • Balm Dotcom
  • Boy Brow
  • Cloud Paint
  • Stretch Concealer
  • Generation G
  • Lash Slick
  • Milky Jelly Cleanser
  • Priming Moisturizer
  • Perfecting Skin Tint
  • Solution 

Given that I wanted to understand consumer engagement and feedback across all popular products, I utilized Pandas, a data analysis library in Python, to tag the reviews with their respective product names and then to merge the reviews together. 

I also noticed that the timestamp scraped from the reviews were in a "string" format. I converted this into a "datetime" format to use in my analysis. 

In some of my analysis, I segment the reviews by consumer age group and by consumer skin type. There were 946 rows that had missing values for either (or both) fields. I decided to omit these missing values when dealing with segmented analysis. 

I also decided focus on the following age groups: 17-24, 25-30, and 31-40. This was due to the limited data availability for the other age groups.

Data Analysis

After a first initial look through the data, I wanted to ask the following questions: 

  • What is the trend in consumer engagement, if any? 
  • How are different subgroups rating the products? 
  • Is there a significant difference in who would recommend the product amongst the different consumer subgroups? 

What is the trend in engagement? 

Although there are many other ways Glossier engages with its consumers, I wanted to focus on the number of reviews left as a metric for engagement. In line with the "democratizing beauty" mantra, providing feedback on purchased products is a way for Glossier consumers to engage with the brand. 

A simple line graph was telling; there is a clear downward trend from mid-2018 to the present in the number of reviews left by consumers.  

Dividing the data by age group or by skin type, it is clear to see that the downward trend from mid 2018 is true for all subgroups. 

How are consumers rating the products?

There is an important factor to consider when observing this data. Since I decided to scrape the most popular products, this may skew how "liked" these products are and should not be generalized for all glossier products.

When observing by skin type, those identifying as the "Normal" skin type seem to have the highest proportion of 1 Star reviews. When slicing the data by age group, there seems to be a trend towards higher proportion of 1 Star reviews as the age group gets older. 

Who is recommending the products?

Given the context that these are reviews of popular products, there seems to be a high proportion of those who replied "Yes - I would recommend this to a friend". But are any of these responses significantly different? i.e. Is there a group that is more likely or less likely to recommend than the others? I decided to approach this question by performing a Chi Square Test of Independence. 

Null Hypothesis: The proportion of "Yes" is the same across all skin types

Alt. Hypothesis: At least one skin type significantly differs in proportion from the rest

Significance Level: 0.05

Result: The resulting p-value of the test was 4.94e-20. At a significance level of 0.05, we reject the null hypothesis. In other words, at least one of the skin type groups differs in their proportion of "Yes" responses. 


In summary, there seems to be a declining trend in user engagement in terms of number of reviews left on Glossier's popular products. When observing the different consumer segments, those with "Normal" skin type seem to be leaving the highest proportion of 1 Star reviews. This proportion of 1 Star reviews seem to increase with higher age groups. 

Future Work

  1. Scrape reviews of all products from and see what makes a product popular versus not popular
  2. Perform a logistic regression to determine which customers are more likely to reply "Yes - I would recommend to a friend" based on skin type, skin shade, age group, and location
  3. Compare the text of 1 Star reviews versus 5 Star reviews to better understand what customers are saying on both sides of the spectrum 

About Author

Sunny Lee

Sunny graduated from Northwestern University with a double major in Economics and Statistics. She joined Goldman Sachs as a Sales Analyst in 2015 and took subsequent roles at the firm as a Fixed Income Trader and as a...
View all posts by Sunny Lee >

Leave a Comment

No comments found.

View Posts by Categories

Our Recent Popular Posts

View Posts by Tags

#python #trainwithnycdsa 2019 2020 Revenue 3-points agriculture air quality airbnb airline alcohol Alex Baransky algorithm alumni Alumni Interview Alumni Reviews Alumni Spotlight alumni story Alumnus ames dataset ames housing dataset apartment rent API Application artist aws bank loans beautiful soup Best Bootcamp Best Data Science 2019 Best Data Science Bootcamp Best Data Science Bootcamp 2020 Best Ranked Big Data Book Launch Book-Signing bootcamp Bootcamp Alumni Bootcamp Prep boston safety Bundles cake recipe California Cancer Research capstone car price Career Career Day citibike classic cars classpass clustering Coding Course Demo Course Report covid 19 credit credit card crime frequency crops D3.js data data analysis Data Analyst data analytics data for tripadvisor reviews data science Data Science Academy Data Science Bootcamp Data science jobs Data Science Reviews Data Scientist Data Scientist Jobs data visualization database Deep Learning Demo Day Discount disney dplyr drug data e-commerce economy employee employee burnout employer networking environment feature engineering Finance Financial Data Science fitness studio Flask flight delay gbm Get Hired ggplot2 googleVis H20 Hadoop hallmark holiday movie happiness healthcare frauds higgs boson Hiring hiring partner events Hiring Partners hotels housing housing data housing predictions housing price hy-vee Income Industry Experts Injuries Instructor Blog Instructor Interview insurance italki Job Job Placement Jobs Jon Krohn JP Morgan Chase Kaggle Kickstarter las vegas airport lasso regression Lead Data Scienctist Lead Data Scientist leaflet league linear regression Logistic Regression machine learning Maps market matplotlib Medical Research Meet the team meetup methal health miami beach movie music Napoli NBA netflix Networking neural network Neural networks New Courses NHL nlp NYC NYC Data Science nyc data science academy NYC Open Data nyc property NYCDSA NYCDSA Alumni Online Online Bootcamp Online Training Open Data painter pandas Part-time performance phoenix pollutants Portfolio Development precision measurement prediction Prework Programming public safety PwC python Python Data Analysis python machine learning python scrapy python web scraping python webscraping Python Workshop R R Data Analysis R language R Programming R Shiny r studio R Visualization R Workshop R-bloggers random forest Ranking recommendation recommendation system regression Remote remote data science bootcamp Scrapy scrapy visualization seaborn seafood type Selenium sentiment analysis sentiment classification Shiny Shiny Dashboard Spark Special Special Summer Sports statistics streaming Student Interview Student Showcase SVM Switchup Tableau teachers team team performance TensorFlow Testimonial tf-idf Top Data Science Bootcamp Top manufacturing companies Transfers tweets twitter videos visualization wallstreet wallstreetbets web scraping Weekend Course What to expect whiskey whiskeyadvocate wildfire word cloud word2vec XGBoost yelp youtube trending ZORI