3 ms·
I don't really understand the methodology on this test. How are you getting a fair control/variant split in Google SERPs?
by ssharp 10y ago
I don't really understand the methodology on this test. How are you getting a fair control/variant split in Google SERPs?
- zck 10y agoThey took 20,000 pages, split them into test and control groups, and changed the title of questions in the test group from "$QUESTION - Coderwall" to "$QUESTION (Example) - Coderwall". They then measured how many more clicks the test group got compared to the control group.