Google Merchandise is planning a new marketing campaign for a DIY kit. Jim Parsons, of The Big Bang Theory, will promote the kit to robotics enthusiasts who want to build a Google-branded twin-rotor drone. The campaign is initially going to run only on social media in three states chosen as test markets. As the leader of the web analytics team at Google Merchandise, you have been asked to choose the three states for the new campaign and identify one social media platform to use in each state. The campaign can run on different social media platforms in each state.

To avoid being misled by seasonal variations, you will need to use the Google Merchandise Demo data (https://analytics.google.com/analytics/web/?utm_source=demoaccount&utm_medium=demoaccount&utm_campaign=demoaccount#/report/visitors-overview/a54516992w87479473p92320289/) from the past year to make your decision. Use the drop-down menu to change the date range; you might also want to use filters to take out states and social media platforms with very few conversions (<20), which should ensure that a few large purchases do not skew your results. Use a filter to exclude states and platforms that have no e-commerce conversions, since you will not be able to use conversion metrics for them.

Then, address the following prompts:

1. Which states will you choose to run the campaign in? Justify your decision using appropriate metrics.
2. For each state, identify one social media platform on which to run the campaign. Why did you choose this platform? Defend your choice by citing and comparing the appropriate metrics.
3. The campaign will be run as paid acquisitions from the selected social media platforms in each state. Identify the maximum cost per acquisition that you will be willing to pay (at this price, you will break even). Assume a GMoS of 12%.

There are no right or wrong answers; your reasoning and how you use the data are what will be considered. Attach screen shots of the custom displays you create, download to Excel for further processing, and attach tables.