I recently did a webinar on driving conversions with Acquia Lift and I wanted to share some of the key details from the webinar with a broader audience. In the webinar I covered my testing strategy, where to start testing, how to u se Acquia Lift, review tests we’ve been doing here at Acquia and see the results we’ve seen. In this blog post I’m going to be focusing on where to start testing and some of the results we’ve seen. To get the full story please watch my webinar.
What to test?
Sometimes deciding where to start with your testing and personalization can be a daunting task. Here is a list of places I’ve started testing on acquia.com that will hopefully give you a jump-start on your own testing.
Button Copy | |
Content on the page | |
Page layout | Source:http://www.ometrics.com/resources/#sthash.sUhR8Joc.dpbs |
Page titles (any test on the page you want) | |
Form submit | |
Button color | |
Menu items | |
HTML | |
Drupal Views vs. Drupal Blocks |
For acquia.com the exciting part about Acquia Lift is the direct integration with my Drupal site because we use the Drupal user registration function I can utilize user profile data into my personalization tests. It will also be fully responsive with my site because its not placed on top, it's built in. You’ll see this more specifically with my homepage banner test, and can also learn more about how we built this test out in Erica’s blog post here: How to: Responsive, Full-width Banners.
Test’s we’ve done:
Asset testing: For this test we have done 2 variations of this test. Test version 1 ran from January 31st – March 4th was then turned off for most of March for module updates and then adjusted and turned back on March 27th – Present. Viewing the results you can see we had significant click through increase in February while the first test ran. We then saw a decrease while no test was running and again an increase in April while the next test ran. These tests were also auto-personalized using our job function and industry user profile options; we gain this data when users select their job function or industry when creating accounts on the site, this way Acquia Lift
can show you relevant information to you. I had chosen assets to display for industries such as government, higher education and media and entertainment along with job functions such as developers or marketers. Here are the results:
Results:
Homepage banners: This test was very exciting and something I wouldn’t have been able to do without Acquia Lift
. I am testing a basic Drupal block against a Drupal views slideshow. This test is fully responsive and tracks homepage user engagement. We have recently turned the test live for everyone but based on these results after only 1 day you can see why. We were seeing similar conversion rates between both variations but this number fluctuate as the test runs. Lift over control and chance to beat control options where quite high based on total test not just at this moment. Lift over control is predicted goal value directly compared to your control, so we have gained a 7.9% goal value by having this new variation. Lift over random takes both variations into account and averages them and tell you how your variations compares. Then you have Chance to beat control, this acts like your confidence reading, we saw a 94% change to beat so I read this as a 94% confidence in my second variation. You can see the data we have below.
Results:
I am continuing to use lift and test new items and learn along the way. I’ve struggled with other A/B testing and personalization tools in the past because of the complexity of my site along with https vs http site structure and responsiveness. So far these hurdles have been reached with Acquia Lift. Please watch the full webinar to gain more insight. Webinar: https://www.acquia.com/resources/acquia-tv/conference/personalization-ho...