(un)Common Logic answers that challenge by examining data to determine which tests would be most effective.

The (un)Common conversion rate optimization testing process starts by using data to determine which areas of your site would benefit most from testing. Then we make a plan. Using the scientific method, we hypothesize what effect each specific change might provoke. We never run a CRO test with “let’s see what happens; you never know” as a premise.

The (un)Common CRO testing plan process has two steps:

1. Create a list of test hypotheses

• Based on data from our CRO audit
• Granular, testing one thing at a time
• A probable outcome established to test against

2. Prioritize hypotheses to test based on several factors

• Projected impact on user behavior
• Financial value of that impact
• Anticipated difficulty of test implementation

The resulting plan determines which tests to run and in what order. We create this plan because we’re not interested in running the most tests. We’re motivated by delivering the most rewards as efficiently as possible.

“Testing without data is just guessing. And testing without a hypothesis is… well, that’s just guessing, too.”

Josh Cuttill, Product Manager