Many of you, frustrated by testing challenges, have vented on this blog and in our optimization group on Linkedin.
Those comments inspired our [July 15] web clinic: Optimization vs. frustration: Overcoming barriers to better tests and gains.
With this clinic, Dr. Flint McGlaughlin and our research team will explore reviewed some of the biggest obstacles to conversion testing and how you can overcome them. To get you thinking, check out these resources related to optimization and testing:
-
- Designing tests for maximum impact (part 1) … Before implementing a testing cycle, there are three factors you need to be aware of that can make a major difference to your outcomes. Andy Mott, our senior research manager, explained these three factors in this post: Better tests, better results – here’s how.
-
- Designing tests for maximum impact (part 2) … Ready to dig deeper? Our team of analysts distilled months of experiments and case studies into our year-end research review, and the overarching theme was planning and executing more successful marketing tests.
-
- Choosing the right pages to test … The Official Google Website Optimizer Blog put together a 60-second video to help you figure out which of your pages most need improvement: Picking a good test page with Google Analytics.
- Testing different stages of your marketing funnel … Jimmy Ellis, our director of optimization, uses a test that achieved a 103% conversion gain to lead you through the key steps of testing, from identifying the right elements to optimizing the offer path, before discussing advanced testing strategies with the Omniture team in this joint webinar and optimization guide: Landing Page Optimization: How You Should Be Testing At All Stages of the Marketing Funnel.
What’s holding you back from testing your way to bigger conversion gains?
I would think the biggest hold back on testing, is keeping variables constant. If the online marketing department is not aligned with other areas of the company specially marketing, then simultaneous offline campaigns might have an effect on the results of the test. This is one scenario, but there are many where the problem isn’t so much conducting the test and setting the stage, whether doing A/B , multivariate etc but just keeping things the same to compare against the benchmak.