Analytics
3K members online now
3K members online now
Dive into multiple domain (Cross/Sub) tracking, implementing Ecommerce and Enhanced Ecommerce, setting up Event tracking, and Universal Analytics code.
 
Guide Me
star_border
Reply

Multiple Content Experiments on a single page? Best practice? Is it a bad idea?

Visitor ✭ ✭ ✭
# 1
Visitor ✭ ✭ ✭

Lets say I have a large site that I want to do a lot of testing on using Content Experiments. Is it a good idea of a bad idea to run multiple tests on the same page?

 

For example, lets say I am running a test for my global navigation in the header. Any page you visit has a header with the global navigation, so it would run on every single page of the site. So now, if there are other tests that I want to run on unique, one-off pages on my site, how do I do this? Because now I would have multiple tests running on my pages: The global nav test and the one-off test.

 

What is the best approach to this problem? Is it a bad idea to try to do this? If so, how do you run tests that are global to the entire site without halting all other testing on the site?

Re: Multiple Content Experiments on a single page? Best practice? Is it a bad idea?

Visitor ✭ ✭ ✭
# 2
Visitor ✭ ✭ ✭
Hello.
If you do not specify whether the test with the aim of any conversion, would everyone hard to answer.
For example, if the Amazon, in order to increase sales, it is always changing button cart in. It is, but only in common with huge content page, because the buttons to the maximum contribution to sales. Then, in the Amazon, the header is surprisingly design'm falling apart. That is, the sales will not contribute much.
Before the AB test, first to a heat map test, after examining whether the user has seen anywhere, but what to do is. Begin if not noticed, or not than effect hardly achieved for a test.
AB test is the farther from the goal, the contribution will be many uncertainties. And get to clarify the goal, it would be easy to answer other people.
Thank you for your attention.

Re: Multiple Content Experiments on a single page? Best practice? Is it a bad idea?

Visitor ✭ ✭ ✭
# 3
Visitor ✭ ✭ ✭
What I did on few occasion was to run one experiment including all the variants, spreading traffic equally. Simplifying it:
original - header1 - picture1 - button1
v1 - header1 - picture2 - button1
v2 - header1 - picture1 - button2
v3 - header1 - picture2 - button2
v4 - header2 - picutre1 - button1
v5 - header2 - picture2 - button1
v6 - header2 - picture1 - button2
v7 - header2 - picture2 - button2

As you can see it can get complicated, so you must have a good system to analyse the results. It also takes obviously more time, unless you have tons of traffic. I usually stick to experimenting one element at a time for a week (100k pageviews) and move forward.