I like the idea of MVT and believe it can work, but how do you know if Headline A was stronger than Headline B if the Calls to action are different and potentially skewing the tesults?

David,

The best thing about MVT is that it will show you the break down of different elements as they interact with one another. When you get the results each item has its own contributed conversion rate, all you normally need to do is mix the highest converting elements. It can also show you where two "high" converting elements might clash and decrease each other´s value. Because each element´s conversion rate is separate from the others, you can quickly see what works and what doesn´t

Ok, so that is Multivariate Testing. Great to know about this.

I can`t understand, how i can calculate results()? Where i can find a formula to calculate the percentage?

Vitaly,

Take the number of people that converted and divide by the number of people that saw the call to action. That will give you the percentage.