To acquire more members for the Official NASCAR Members Club, we worked with NASCAR and their own database of fans to promote our membership. Unaware of the performance of their list, we created factorial design experiments to optimize the results.
Using a 95% confidence interval and 2% margin of error, we created random sample sizes of 5,000 per send. While we could determine the best subject line and landing page separately, we wanted to see how all these variables interacted with each other so we can send the best possible combination. Ultimately, the goal of a factorial design experiment is to test multiple factors and interactions at once to optimize a campaign as quickly and inexpensively as possible.
Increase acquisition while obtaining the best click, open and conversion rate.
Designed and implemented experiments testing 8 combinations of three separate factors (Subject Line, Email Template, Landing Page Presentation). This allowed us to test the interactions between these three factors together in one controlled experiment to 8 random samples of 5,000. The winning combination from the experiment, allowed us to optimize the remaining bulk of the send, while providing us with an educated projection for the performance of this large full send. A second chance email was created and launched to those who did not take advantage of the original offer.
Tested a campaign with multiple testing variations and launched a final version in a week.
Using the experiment featuring the different combinations, allowed us to optimize the final send, providing us with the best conversion rates possible.
May 14, 2020
Latest Blog Posts
April 22, 2020
3 ways to easily leverage Machine Learning in your Facebook and Google campaigns
Machine Learning (ML) may sound like a futuristic and complex concept to implement in our advertising campaigns; however, Facebook and Google already have some powerful ML techniques that you can...