Schedule 2024

Geo experimentation has been a mainstay in marketing measurement for quite some time now. With privacy restrictions becoming stricter, geo experiments are likely to only gain in popularity. However geotesting tends to have a reputation of being overly complex, expensive and broad.
Picture this: you've finally convinced your team that we should use an experiment to validate that last-click attribution model they want to use to value SEM Non-Brand. You secure the budget, set up a geo experiment, run a test for 6 weeks, and then…there's no detectable lift because there was noise in the pre-period. What now?
In this talk we'll explore how Etsy has improved our geotesting methodology and how we're reaching more precise results in shorter time frames. We'll share anecdotes about how specific tests have influenced our processes, how to get started with geo experimentation, open source resources, and suggestions on how to improve your geotesting precision.

Takeaway

• What is geo experimentation and how is it used?
• Overview of the existing challenges and opportunities in geo experimentation

• Methods on how to increase precision of experiment results
• Survey of open source geo experimentation resources