Rich-text Reply

Running an experiment on an element that only appears sometimes

danielhochman 10-22-15

Running an experiment on an element that only appears sometimes

When running an experiment on a marketing unit/elelement which only appears on some occurances of the page and not others, what does optimizely count for usage purposes (MUVs)?

 

For example, all pages have the same url structure site.com/article/2323 , but some articles have this marketing unit / element and some do not. Does it count every view of site.com/article/ or only the views where the element being tested appeared?

 

 

David_Orr 10-22-15
 

Re: Running an experiment on an element that only appears sometimes

[ Edited ]

Hi Daniel,

 

That is a great question. An MUV is counted if a visitor is able to get into an experiment at least once in the current month. If the experiment continues to the next month, the same visitor's will be counted towards the MUV allotment even if they do not activate the experiment. The reason is because their conversions will still be attributed to an experiment as long as the experiment is active.

 

Here is an example:

 

Month 1: Visitor is bucketed into the experiment. This visitor will be counted towards the current month's MUV.
Month 2: The same visitor from month 1 visits the site where Optimizely is enabled. This visitor will be counted towards the current month's MUV. All subsequent visits in this month will not change the MUV since MUV tracks unique visitors rather than page views or sessions.

 

Let me know if you have any questions.

 

David

Senior Technical Support Engineer
Optimizely

Re: Running an experiment on an element that only appears sometimes

Thanks David.

I appreciate the response, but this didn't really answer my question. I wasn't asking about month to month tracking for MUVs. I was asking about a scenario where I'm targeting a specific page for an experiment. The experiment is changing text. However, this text only appears on the page 10% of the time it is viewed.

My question is as follows- what does optimizely count towards MUVs? The 10% of the page views where the text is shown and changed by optimizely or 100% of those PVs since optimizely tried to find that text and didn't. If the latter, how can I target the experiment to only run when the text appears so I only need to pay for the PVs where the experiment actually changes something?