P.S. Also, does your software's MVT optimize itself like genetify - http://www.shoemoney.com/2010/03/03/split-testing-with-a-gen...
P.S. Also, does your software's MVT optimize itself like genetify - http://www.shoemoney.com/2010/03/03/split-testing-with-a-gen...
No, VWO doesn't dynamically optimize because statistics become s really complicated once you try to model such optimization. As far as I know Genetify (or any other tool out there) doesn't have a statistically valid method of optimizing the test (while it is running).
The math required to do such optimization while maintaining statistical validity is beyond me, at the moment. (Though I am actively researching on this front. Any pointers/links will be appreciated)
Genetify is open source - http://wiki.github.com/gregdingle/genetify/ - you can have your programmer go through it for the algorithm. That'd be a good starting point.
Statistically valid optimization would be a HUGE plus. Enough for me to pay for VWO instead of using genetify.
Third party cookies (via IFRAME) is a choice, it isn't enabled by default for all tests.