Junduo123 / Analysis-of-Effectiveness-of-branded-paid-search.

This is the 2019 MMA Datathon project related to conversion of web traffic analysis. Result: 3rd place finalist.

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Competition:

2019 University of Toronto Master of Management Analytics Data Hackthon

Challenge:

Analysis of Effectiveness of branded paid search

Background

  • By 2018, dozens of competitiors in SOBSS (Simple On-Line Business Solutions) space when people searching for "business solutions for small business". Attracting new customers became both harder and more expensive.
  • A cloud-based Software as a service (SaaS) company provides a variety of basic business software. Major clinets of this company are small and medium-sized companies in the US and Canada.

Problem

  • CEO and CMO can not reach an agreement about whether continue to invest in Branded PS in the next Fiscal Year
  • CMO insists that Branded PS is the most efficient marketing channel
  • CEO, however, argues that marketing spend on Branded PS is misguided and he predicts when Branded PS turn off customers will find the website through organic search, therefore spending should be redirected to Non-Branded PS or other channels

Presentation Slides: This is our presentation slides outlining our analytical roadmap, findings, recommendations, along with appendix: Presentation Slides

Useful Code: We use R language as our analytical tool to conduct statistical test for this challenge: Code.Rmd

Our Team's submission to the Datathon held at University of Toronto, Oct 9 2019. Third place Finalist

About

This is the 2019 MMA Datathon project related to conversion of web traffic analysis. Result: 3rd place finalist.