We  just completed two A/B tests on two different websites to test out some elements on a homepage to see which one makes more money.

#1 was testing putting a quick search higher on the page (right side) Vs. lower on the homepage.  The % of visitors who searched inventory went up a whopping 11.2%.  This test ran for a few weeks and achieved a "certaintly" level of 99.2%

#2 was to test a pop-up email registration form (you can see this if you go to my website at www.blizzardinternet.com and view a few pages).  The theory was that putting an email signin popup would **** people off and they would bounce!  During a 42 day test, we actually found that a visitor was 7.2% more likely to view a second page IF we showed them a popup asking for email. Crazy!  So, we not only get hundreds and hundreds of email addresses, our bounce rate went DOWN!

Have you ever tried an A/B test??