A/B-тесты прелендингов
Создание прелендингов · 4 мин чтения
Why A/B Test?
Even small changes to headlines, images, or button text can dramatically impact conversion rates. GramPixel's built-in A/B testing lets you compare variants with statistical significance.
Setting Up a Test
- Go to A/B Tests in the sidebar.
- Click Create Test.
- Select the pre-landing pages you want to compare (up to 5 variants).
- Set the traffic split (e.g., 50/50 or 70/30).
- Define the winning metric — join rate, click-through rate, or custom conversion.
Reading Results
- Visitors — total unique visitors per variant.
- Conversions — users who completed the target action.
- Conversion Rate — percentage of visitors who converted.
- Statistical Significance — GramPixel calculates confidence levels automatically. Wait for 95%+ confidence before declaring a winner.
Best Practices
- Test one variable at a time (headline, image, CTA button).
- Run tests for at least 7 days or 1,000 visitors per variant.
- Don't stop tests early — let them reach statistical significance.