Closed garethrees closed 8 years ago
Before (6/2/16 - 8/3/16)
After (10/3/16 - 10/4/16)
IIRC the methodology was:
The date ranges look odd but should represent the same number of days. The skipped day is the day the change was deployed so was deliberately excluded due to the uncertainty of which version of the site was being reported.
Unfortunately the Goal Conversion data is unsuitable because there is not complete data available for the before period so the comparison will be misleading.
Is it worth doing this same experiment on AskYourGov? They currently make users log-in before making a request. This would obviously require getting access to their Google Analytics beforehand though and setting up goal to track how many people make requests.
No - not worth experimenting again. Let's just take it off their site.
Q: Does the requirement to log in before making a request decrease conversion?
[Explain how you will measure this]
Measure conversion rate Don't have goals enabled but can look at /request/:url_name/new
[What do we need to do to be able to take the measurements?]