TRI MSDM AO #4 Overview

Table of Contents:

• Introduction

  • Problem Statement

  • Analytical Objective #4

  • AO Importance

• Literature Review

• Methods

• References

Introduction

• Problem Statement
  • TRI faces high bounce rates and low conversion rates despite frequent site visitors.
• Analytical Objective #4:
  • Identify ways to improve conversion rates for training bookings, donations, and newsletter signups.
• Importance:
  • Higher conversion rates may lead to an influx and increase in bookings or donations, allowing TRI to grow and impact other teams through training resources, as well as receive funding for important and impactful projects.

Literature Review

• Existence of well-established consumer behavior model:

  • Minimal Cognitive Process Theory (MCPT): This theory implies that there are generally two variables that attribute a user staying or leaving a website.

    • Clickstream Cost

    • Personal Cost limit

  • The more complicated a website, the lesser chance of a successful purchase.

  • The longer it takes to convert, the greater the chance a visitor leaves.

Literature Review Cont.

• Assessment of the adequateness of gathered information:

  • Website optimization is essential & so is ease of navigability.

  • Visitors who come to the site have an intent.

  • Nuances between for profit & non profit can vary in messaging.

• Is there additional analysis needed to address the AO?

  • GA4 data that is being accessed and assessed via BigQuery, and then translated into Looker Studio for robust dashboards and metrics.

Methods

Data Sampling:

• Population: TRI site visitors, clinicians, donors, and others who may interact with TRI’s digital ecosystem.

• Sampling: Email CTR, page views, bounce rates, session duration metrics, clicks, impressions

• Data Wrangling: See below steps:

  1. First, we will need to gather the correct data from GA4 and assess via BigQuery & R.
  2. Secondly, we will need to transform the data using BigQuery and SQL.
  3. Lastly, we will need to visualize via Looker Studio or Tableau.

Methods Cont.

• Sampling Characteristics:

  • Sample characteristics could include:

    • Active users 18-24 or 25-44

    • Gender: 75% Female, 25% Male

    • Familiarity with mental health, wellness and traumatic experiences support.

• Measures:

  • Independent Variables: Google Ads campaign, conversion pathway, landing pages

  • Dependent Variables: Engagement metrics to Google Ads, user engagement, bounce/conversion rates

• Analytics Methods: Google Analytics Explore, Logistic Regression Random Forest

  • These methods will be used to create funnel exploration to map user behavior, and identify user signals.

References:

• Gkikas, D. C., and P. K. Theodoridis. 2024. “Predicting Online Shopping Behavior: Using Machine Learning and Google Analytics to Classify User Engagement.” Applied Sciences 14 (23): 11403. https://doi.org/10.3390/app142311403.

• Melinevskyi, A., S. Koberniuk, T. Bilousko, V. Vasiuta, and N. Strochenko. 2023. “Digital Marketing and Its Role in Customer Acquisition.” Economic Affairs 68 (4): 2229–2238. https://doi.org/10.46852/0424-2513.4.2023.31.