Googlin Online Merch Store Performance: Analysis of Google Merchandise Store

Avatar
Posted on Oct 26, 2020

Background

It is possible that you didn’t know about it, but Google has an official online merchandise store where they sell their branded products and those of their other popular brands such as Android, YouTube, and so on. The question is, is there really a market to sell Google-branded products such as t-shirts, cups, stickers, and so on? Thanks to the Google Analytics team, these questions can be answered by looking at Google’s real e-commerce data that became partially available. However, we don’t want to simply just observe the data and answer some conventional e-commerce marketing questions. We want to further provide actionable insights and recommendations based on the available data. Hence, by creating an imaginary scenario in which Google is trying to expand its online merchandise store, this project will walk you through Google’s merchandise sore data and provide insights based on the analysis by tackling the following problem statement:

How can Google online merchandise increase its transaction number by attracting more consumers, improving the online store’s user experience, and coming up with better strategic planning based on analyzing seasonality patterns for the next calendar year?

About the Data

A big thanks to the Google Analytics team, Google merchandise store is publicly accessible through Google BigQuery, a web service that lets developers and analysts conduct an analysis of big data sets. Unfortunately, because the available data is a demo version, it only had data from August 1, 2016, to July 31, 2017. Luckily, much important information you might find on a conventional e-commerce website is included such as the following information: traffic data, content data, and transaction data.  

  • Traffic data includes information about where online merchandise store users originate and how they interact with the website. This includes information about organic traffic, paid traffic, time spent on the website, hit numbers, and bounce rate.
  • Content data includes information about the behavior of the users on the site, which includes the URLs of pages that users look at, and how users interact with the page contents through bounce rates, hit numbers, and time spent on the website.
  • Transaction data contains information about the transactions that occur on the Google merchandise store. This includes transaction numbers and revenue numbers.

In order to make the most out of the demo version, I queried the data to contain most of the important variables that were available between August 2016 to July 2017,  which resulted in a total of 901,097 visits and 12,070 transactions record. 

Part 1: How well is Google Merchandise Store Doing?

So how well is Google merchandise store doing across the year in terms of transaction numbers? Based on the graph above, October and November in 2016 have high transaction numbers that we should be aware of. November, in particular, has 50 % more transaction numbers than any other month.

When we observe the transaction number by country, our consumers are mostly purchasing from the United States, indicating the United States as Google merchandise store's biggest market. In terms of the product by category, apparel was the most popular merchandised product category. 

So back to our original question - how well is Google online merchandise store doing? An explicit answer to that question is we don't know since we don't have any information on Google merchandise store's previous performance. However, our key takeaway from this simple transaction data overview are the followings: 

  • One, compared to other months, November has a high volume of transaction numbers, directing possible abnormality in terms of analyzing seasonality patterns. Therefore, Google merchandise store can come up with improved strategic planning to target new or returning consumers that visit the website during November.
  • Two, the United States is the biggest market for Google merchandise store. This means that there are more opportunities beyond the United States that Google can focus on in order to expand its market.
  • Three, among the many merchandised products Google is selling, apparel comes out on top. Hence, expanding Google's apparel products by giving more options and more products to buy can lead to more transaction numbers in the future.

Part 2: How Long Does Our Consumer Stay on Our Website?

So far we have looked at some basic summary statistics on Google merchandise store performance based on transaction numbers. However, the real game-changer in e-commerce analytics is around traffic data and content data. It is obvious that every marketing and promotional decision is driven by data. Without data, you don't know what is working, what is failing, or what success even looks like. E-commerce is no different. In fact, making decisions and driving growth in e-commerce requires more in-depth analysis such as tracking where your users are coming from and monitoring your users' activities besides setting key performance indicators, identifying necessary metrics, and making adjustments when necessary.

One of the key concepts of e-commerce analytics is funnels. The idea is simple - your target user will go through a step-by-step flow until the user makes a transaction, just like a funnel. For example, 1) your user is looking for a particular product and happens to find your website thanks to Google Ad. 2) Your user lands on your home page and your user is satisfied with the result and decides to look around. 3) Your user finds a satisfying product and clicks on 'add to cart'. 4) Your user clicks to check out. 5) Your user enters their personal information and finalizes the transaction. At each step, a certain percentage of people might drop out of the funnel. Knowing such percentage or knowing when they drop out can help you determine where your website is lacking and understand the psychology behind your users.

Based on this knowledge, we will analyze the traffic and content data around Google merchandise store and see how users are interacting with the online store by looking at the following variables: hit numbers, time on site, bounce rate, channels, devices, and so on.

In terms of traffic data, all traffic can be divided into several categories - organic search, referral, social, direct, paid search, and so on. Based on the graph above, Google merchandise store is visited mostly via organic search, which means most of our users are coming through unpaid search. In other words, users visiting our website from a search engine's organic results have a very specific intent and if we can provide them with a satisfying solution, these users are more likely to convert. 

 

Next, if look at our number of visits by devices, most of our users are using desktops to access our online merchandise store. This is most likely going to be the case since we currently don’t have any apps available. However, given the growing importance of mobile devices in the e-commerce scene, it is crucial to expand different options in devices for users to experience Google merchandise store.

When we look at the bounce rate, our home page has a bounce rate of approximately 30%. As an online merchandise store. We want our consumers to view more than one page, so having a high bounce rate is bad. In order to have a lower bounce rate, it is important to enhance our user experience with our online store.

Based on the graph above, having a greater number of hits is correlated with greater revenue.

Limits

Because our dataset only included dates only from 2016.08.01 to 2017.07.31, it is difficult to make a solid analysis. If possible, investigating a few years of data should be necessary in order to analyze seasonality patterns.

Also, because the BigQuery data was a demo version, many important variables were either missing or were difficult to access. Data regarding hits and ads were too large (20GB +) for R to handle, so it was difficult to query and other important data such as images related data (absence of image or not in a specific page) were not available to the public.

Future Work

In the future, I would like to answer the following questions: Are there any abnormally high volume of traffic or transactions before thanksgiving? If so, how many days? This can be important when it comes to strategic planning.

About Author

Leave a Comment

No comments found.

View Posts by Categories


Our Recent Popular Posts


View Posts by Tags

#python #trainwithnycdsa 2019 airbnb Alex Baransky alumni Alumni Interview Alumni Reviews Alumni Spotlight alumni story Alumnus API Application artist aws 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 Bundles California Cancer Research capstone Career Career Day citibike clustering Coding Course Demo Course Report D3.js data Data Analyst data science Data Science Academy Data Science Bootcamp Data science jobs Data Science Reviews Data Scientist Data Scientist Jobs data visualization Deep Learning Demo Day Discount dplyr employer networking feature engineering Finance Financial Data Science Flask gbm Get Hired ggplot2 googleVis Hadoop higgs boson Hiring hiring partner events Hiring Partners Industry Experts Instructor Blog Instructor Interview Job Job Placement Jobs Jon Krohn JP Morgan Chase Kaggle Kickstarter lasso regression Lead Data Scienctist Lead Data Scientist leaflet linear regression Logistic Regression machine learning Maps matplotlib Medical Research Meet the team meetup Networking neural network Neural networks New Courses nlp NYC NYC Data Science nyc data science academy NYC Open Data NYCDSA NYCDSA Alumni Online Online Bootcamp Online Training Open Data painter pandas Part-time Portfolio Development prediction Prework Programming PwC python python machine learning python scrapy python web scraping python webscraping Python Workshop R 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 Selenium sentiment analysis Shiny Shiny Dashboard Spark Special Special Summer Sports statistics streaming Student Interview Student Showcase SVM Switchup Tableau team TensorFlow Testimonial tf-idf Top Data Science Bootcamp twitter visualization web scraping Weekend Course What to expect word cloud word2vec XGBoost yelp