Bucket test
A bucket test is a split test or A/B test in which you compare two versions of a website against each other, to determine which one is more effective in generating conversions – purchases, downloads, subscribers, clicks and so on. One of the pages is considered the control, and the other one is a variation, visitors being exposed to one of the variations.
For each bucket test, you begin with a hypothesis – a text, design or usability change that you think will be effective in improving conversions. You test that hypothesis and if the variation performs better than the control page for the key metrics, you replace the control page with the variation on your website.
You can test several hypotheses for a page, or do several bucket tests until you’re happy with the conversion rate of that page. Results should be collected until you reach a statistically significant difference in conversions.
'Computer Science' 카테고리의 다른 글
[AI/ML] robust하다 라는 뜻은 무엇일까? (0) | 2024.04.13 |
---|---|
[docker] 사용하지 않은 image 삭제하기 (0) | 2021.09.18 |
은근히 모르는 사람이 많은 내용(나만일지도?) (0) | 2019.03.27 |
Virtual Box 네트워크 설정 (0) | 2014.12.03 |
IEEE Journal, Proceeding, Magazine, Transaction, Letter의 차이 (2) | 2014.06.30 |