A/B Testing

NOTE: This file has some general information. Please read the Mac/Windows specific readme file for specific instructions.

Overview

A/B testing is a user research technique, which is used to test whether the A or B version is a better design.

For a commercial marketplac, the success metric could be calculated based on "conversion rate".

What does this project do?

This project does simple A/B testing of 2 versions of a website. The metrics collected could then be used to see which version among the baseline(A) or variant(B) performed better.

For detailed instructions on how to set it up, please see the README_Mac.md or README_windows.md in this folder.

Sample output after hosting in heroku.

2017-09-22T04:15:05.318219+00:00 app[web.1]: AB_TESTING: A 1506053696735 1506053705304 mp1 1506053696475 2017-09-22T04:15:07.288130+00:00 app[web.1]: AB_TESTING: A 1506053696735 1506053707279 mp2 1506053696475 2017-09-22T04:15:08.353015+00:00 app[web.1]: AB_TESTING: A 1506053696735 1506053708343 mp2 1506053696475 2017-09-22T04:15:09.444328+00:00 app[web.1]: AB_TESTING: A 1506053696735 1506053709431 mp1 1506053696475 2017-09-22T04:15:14.822574+00:00 app[web.1]: AB_TESTING: B 1506053712624 1506053714784 ca1 1506053712526 2017-09-22T04:15:16.682459+00:00 app[web.1]: AB_TESTING: B 1506053712624 1506053716671 ca2 1506053712526 2017-09-22T04:15:17.769221+00:00 app[web.1]: AB_TESTING: B 1506053712624 1506053717759 ca1 1506053712526