Analyzing Information and Transaction Data Using Google Analytics Assignment

Assignment Task

Section A – Analyzing information using Google Analytics

Question 1

In the Acquisition > All Traffic > Source/Medium report, when you add a secondary dimension of “Sales Region,” what are the bounce rates for the keyword “Google Merchandise Store” in North America and EMEA (Audience: All Users, Date Range: 01 January 2021 to 31 March 2021)?

Question 2

In the Audience > Mobile > Overview report, create two Custom Segments - one for male users aged 18-24, and one for female users aged 18-24. Of these two segments, which group had the higher pages/session for mobile and tablet devices (Date Range: 01 January 2021 to 31 March 2021)?

Question 3

In the Audience > Benchmarking > Channels report, select an Industry vertical of “Shopping > Apparel > All Apparel,” a Country/Region of “India > All Regions,” and daily session size of “1000-4999.” Compared to the benchmark, what are the average session durations for the following default channel groupings: Organic Search, Referral and Social (Audience: All Users, Date Range: 01 January 2021 to 31 March 2021)?

Question 4

In the Audience > Overview report, create two System Segments: New Users and Returning Users. For these two segments, report a comprehensive overview of: Sessions, Number of sessions per user, Pageviews, Pages/session, Average Session Duration and Bounce Rate. Compare these metrics with the same period in the last year, that is, 2020 (Date Range: 01 January 2021 to 31 March 2021, compared with the Date Range: 01 January 2020 to 31 March 2020). Use the following format to fill in your answers:

Section B – Analyzing transactional data for association rules

A large supermarket wants to promote two new brands (one soda brand and one yogurt brand) to be launched next week to its existing customers. It wants to suggest additional products as recommendations to customers who are most likely to purchase these two brands. To generate these recommendations a historical database of purchase data for 38,765 transactions conducted by 3,898 customers is presented in the file Groceries_dataset.csv (data source: https://www.kaggle.com/heeraldedhia/groceries-dataset). The KNIME workflow to conduct Market Basket Analysis and generate association rules is presented in the file MRA-Assessment2-Assoc.knwf. Using this workflow and the data set (variable description presented below) answer questions 5 and 6

Question 5

Using a Minimum Support of 0.01 and Minimum Confidence of 0.4 generate 5 association rules for soda. Explain why the top two (i.e., top two with maximum lift) among these rules are also good choices as purchase recommendations when a customer has added soda into their basket

Question 6

Using a Minimum Support of 0.1 and Minimum Confidence of 0.2 generate recommendations for yogurt. Provide qualitative arguments for/against the recommendations that form the output of the Market Basket Analysis.

Section C – Predicting churn and order value from transactional data

Data in this section is extracted from a telecom company that served 7,043 customers in California. Apart from data on whether a customer churned or not, there is data on several other variables (variable definitions are presented in the table below).

Question 7

What are the top three reasons why customers churn (use the Value Counter node in KNIME)? Based on this analysis, what suggestions will you give the telecom company to improve retention rates?

Question 8

(a) Report all pairwise correlations between the variables: Tenure Months, Monthly Charges and CLTV

(b) Using your understanding of CLTV and the values reported in part a, explain why the observed correlation coefficient for Tenure Months and CLTV is positive

Question 9

(a) Estimate a linear regression model with CLTV as the outcome/dependent variable (i.e., Y) and the following variables as independent variables (i.e., X variables): Gender, Tenure Months, and Monthly Charges. Report the statistical significance of all the independent variables. The output should be in the following format:

This Data Analysis has been solved by our PhD Experts at My Uni Paper.