Title: How A/B Testing Drives Better Business Decisions in Digital Media
In the digital age, data is king. Businesses face daily decisions that impact user engagement and revenue. A/B Testing, a tried-and-true statistical method, offers a way to test different strategies and pick the most effective one.
This project explores how statistical testing helped determine whether replacing Public Service Announcements (PSAs) with advertising efforts would improve user engagement. Using a dataset of 20,000 customers, we calculated conversion rates for both control (PSA) and treatment (Ad) groups.
Here are the key results:
The control group had a conversion rate of 3.23% (95% CI: 2.84% - 3.62%).
The treatment group achieved a conversion rate of 6.66% (95% CI: 6.22% - 7.11%).
The lift was 106.01%, and the Z-test p-value was effectively 0, confirming statistical significance.
The effect size, Cohen's h = 0.16, suggests a modest but meaningful improvement.
These results demonstrate how data-driven decisions can guide businesses to strategies that maximize engagement and revenue. This project highlights the practical application of A/B Testing in real-world scenarios, showcasing the power of combining statistical rigor with actionable insights.
For more details, check out the project's full Github Repository.
