Hi, thanks a lot for this video, i've been looking for it for a while. I still have a question, the control and variant are totally different pages, so how you decide which one matches with which one ?
Hey! So when I set a control group, they need to match the variant. This is so no extra variables are added. For example, if I want to test something on my blog, I would pull a group of blog pages and separate that into two groups. Hope that helps!
I still have questions. Why is that we need two sets of pages- to reduce noise and bias from one page? What is the actual treatment. In your treatment group, you rename all 3 titles for all three treatment pages. Does the result show this set of changes are better than changed before? What if two pages title changes are successful but one page title is worse? The actual treatment is vague because it all three pages titles changed differently.
Hi, thanks a lot for this video, i've been looking for it for a while. I still have a question, the control and variant are totally different pages, so how you decide which one matches with which one ?
Hi! How do you select your control? any relation to your experiments? do you also need to set them up in a certain way? Thank you!
Hey!
So when I set a control group, they need to match the variant. This is so no extra variables are added.
For example, if I want to test something on my blog, I would pull a group of blog pages and separate that into two groups. Hope that helps!
@@smamarketing Thank you for answering so quick!
Interesting. Thanks for sharing :)
You bet!
I still have questions. Why is that we need two sets of pages- to reduce noise and bias from one page?
What is the actual treatment. In your treatment group, you rename all 3 titles for all three treatment pages. Does the result show this set of changes are better than changed before? What if two pages title changes are successful but one page title is worse?
The actual treatment is vague because it all three pages titles changed differently.
Why we won't use the Google a/b optimizer tool bro? And how to calculate the forecast for the ppc like this bro
Google optimize is used for CRO. As I shared in the video, SEO a/b testing used a unique approach.
I’ll expand on this tiki pic in future videos.
A/B testing works if you have a massive audience. Don't try to provide it with a small audience.
With the advancements like Casual Impact, even smaller audiences can run effective A/B testing for SEO.