Conversion Rate Optimization
Conversion rate optimization (CRO) is the process of optimizing your paid search ads, landing pages, and overall website design to raise your conversion rate. In other words, the goal is for the highest possible percentage of visitors to your site to convert, or complete your desired action without raising your advertising spend.
CRO Process:
CRO begins by first identifying what the conversion goals are for any given web page or app screen. The success metrics of your website or mobile app will depend on the type of business you’re in, and what your goals are. From here, most companies like to jump straight into testing new designs, and variations of their pages. I prefer to move into a quick "problem analysis" session to determine what specific areas or features should be optimized.
From there it's time to set a hypothesis on what you plan to change, and what you think the result of that change will be. It is critical to set a hypothesis and descriptive statement about the goals of any CRO testing before you being as doing so will help you track your optimization efforts and prevent you from making far too many adjustments in any one test - which can lead to false learning.
Optimization & Testing Platforms:
One of my favorite testing and conversion rate optimization tools is Visual Web Optimizer (VWO). The business package allows you to establish multi-variate tests, build conversion rate funnels, and add heat mapping to your landing pages. Working in concert these tools allow you to understand your user's behavior more deeply than simple analytics reports and provide real insight into your UX UI infrastructure and how optimizations drive behavioral changes.
The key to testing with confidence lies in the proper implementation of the tracking codes, and a thorough declaration of goals to be met along your conversion funnel. With these items in place, the testing platform will randomly send traffic to whatever user segmentation you design and show the statistical significance that one or more variations of your site element beating the "control" or un-optimized group of users.
I find it useful to let most tests run until they hit a Bayesian statistical significance of 90% or higher in terms of conversion rate, but often run tests for quick learnings when tests are color theory or price sensitivity analysis.
Conversion Rate Optimization Software Mastery
Visual Web Optimizer
Optimizely
Crazy Egg
Omniconvert
Google Optimize
HotJar
