Landing Page A/B Testing

Insights and Findings from the A/B Test Analysis

1. Key Observations from the Data

Conversion Rates:

The old page had a conversion rate of ~12.04%.

The new page had a slightly lower conversion rate of ~11.88%.

The observed difference was -0.0016 (0.16%), favoring the old page.

Statistical Significance:

The p-value (0.905) far exceeded the Type I error threshold of 0.05.

The z-score (-1.31) indicated no significant difference between the two pages.

Conclusion: The difference in conversion rates is not statistically significant.

Country Impact:

Added dummy variables for countries (UK, US, CA).

No significant effect of country on conversion rates (p-values > 0.05 for all country interactions).

2. Hypothesis Testing Results

Null Hypothesis (H₀): The new page’s conversion rate is ≤ the old page’s (p_new - p_old ≤ 0).

Alternative Hypothesis (H₁): The new page’s conversion rate is > the old page’s (p_new - p_old > 0).

Outcome: Failed to reject H₀ due to insufficient evidence (high p-value).

3. Practical Implications

No Business Justification to implement the new page, as it did not improve conversions.

Potential Next Steps:

Test other page variations (e.g., layout, call-to-action).

Investigate user behavior metrics (e.g., time on page, bounce rates).

Extend the test duration to gather more data if needed.