From the course: Marketing Attribution and Mix Modeling

Unlock the full course today

Join today to access over 23,200 courses taught by industry experts.

Geo and lift testing to prove incrementality

Geo and lift testing to prove incrementality

From the course: Marketing Attribution and Mix Modeling

Start my 1-month free trial Buy for my team

Geo and lift testing to prove incrementality

“

- [Instructor] A/B tests are ideal for proving what's driving performance. However, in some channels, it's not easy and may even be impossible to run a true A/B test. Geo tests and lift studies are often used to work around these limitations to measure what would have happened without your change and calculate the true incremental uplift. A geo test is the same as an A/B test, except you're randomly assigning geographies instead of assigning users. So for example, if you split the United States into two different groups, you have 25 states in each, and these are randomly assigned, then you can go ahead and make your change in test group B and see how it compares to the control test group A. It's good to use geo testing when it's difficult to A/B test. So for example, with SEO, TV ads or new product launches, you might want to split by geo region or by country if there's a further limitation on regions. It's only useful to…

Contents