Google / MRS: The Essentials of Experimental Marketing vol 2 - Experiment design and analysis

This session, the second of a three-part series, focuses on using experiments to prove marketing effectiveness. In this module we cover Google’s geo experiment methodology and R package to design and analyse experiments.

Experimentation is an integral part of Google’s DNA. We started as an experiment on ‘organising the world’s information and make it universally accessible and useful’, and today, internally run over 7000 experiments annually. We also help advertisers apply scientific discipline and rigour to testing new marketing strategies. We have several self-served testing tools and products that both publishers and marketers can use. This session, the second of a three-part series, focuses on using experiments to prove marketing effectiveness. In this module we’ll cover Google’s geo experiment methodology and R package to design and analyse experiments.

Who will find it useful?

Analysts and Data Scientists working on media evaluation and comfortable with statistical inference using R. Attending the previous or following modules is recommended but not mandatory.

Objectives

• Understand differences and applications of different experiment frameworks
• Review Google’s proprietary Geo Based Regression (GBR) and Time Based Regression (TBR) models
• Design and analyse an experiment using Google’s Geo Experiment package