Can we run A/B test for additional days to achieve sample size(visitors)?

Avatar

Avatar
Level 1
Sundhara
Level 1

Likes

0 likes

Total Posts

1 post

Correct reply

0 solutions
View profile

Avatar
Level 1
Sundhara
Level 1

Likes

0 likes

Total Posts

1 post

Correct reply

0 solutions
View profile
Sundhara
Level 1

14-06-2021

I have an A/B test run on my client site, but it has not achieved the significant number of visitors within the estimated time period for the test. Can anyone confirm if we could run the test for more number of days to achieve the significant/estimated number of sample size? If so, how many number of maximum days we could run the test further? And can we able to get the significant test results with running it for additional days than the initial estimation?

Thanks in Advance!

Accepted Solutions (0)

Answers (1)

Answers (1)

Avatar

Avatar
Give Back 25
Employee
ryanr701
Employee

Likes

203 likes

Total Posts

224 posts

Correct reply

91 solutions
Top badges earned
Give Back 25
Give Back 10
Give Back 5
Give Back 3
Give Back
View profile

Avatar
Give Back 25
Employee
ryanr701
Employee

Likes

203 likes

Total Posts

224 posts

Correct reply

91 solutions
Top badges earned
Give Back 25
Give Back 10
Give Back 5
Give Back 3
Give Back
View profile
ryanr701
Employee

14-06-2021

Hi @Sundhara,

I presume you used the test-duration calculator to determine how long the test should be run, and you are wondering if statistically speaking if letting it run longer is going to compromise the statistical calculations. I think the "days" measure is the less important measure of duration. The more important ones from a stats perspective are total visitors and total conversions per experience. So in short, I'd say yes, run it until you reach the pre-determined per experience visitor volume you need to get a reliable reading on confidence level.

Hope that helps!