I've designed new site structures for one of our clients, and want to test which brings better conversion rates.
- using A/B testing seems obvious but are there any other solutions to cite differences?
Conversion rates are something you can only test with users. However, functionality and other usability issues can be analyzed without users. For example:
If, all other things being equal, one design forces more steps on the user, the barrier of entry is too high and you will have lower conversion rates. Also, if the important buttons are prominent on one design, but not on another you will have lower conversion rates. Big red glowing buttons are hard to resist--but might overpower your design. You'll be able to at least predict what you are going to see before you do the testing.
If, all things being equal (how noticeable the buttons are, and the number of steps needed), you'll find that it really depends on your audience. People are notoriously hard to predict some times.
There are endless tools you can use to increase conversions, it's a whole world that like many other things (SEO for example), that has a lot to do with UX design, but often requires a separate set of skills that might have more to do with marketing.
To name a few tools that you should check out:
Check out this blog post for more ideas: http://www.conversion-rate-experts.com/articles/101-google-website-optimizer-tips/