Prioritize conversion tests
1. Look at your existing conversion funnel data to understand where leaks and drop-offs are occurring.
Break the overall conversion into a series of smaller steps to identify, plan, and implement incremental improvements to the funnel one step at a time.
2. Prioritize meaningful tests by first understanding where gaps actually exist based on real data, and how much time and traffic statistically significant outcomes will take.
Use real data – not guesses or perceptions. Try checking your analytics to get a sense of how long it will take for you to get at least 100 conversions.
3. Use CXL’s PXL framework to make your ratings more objective and foster a data-informed culture.
Make a copy of the available spreadsheet, grade variables using a binary scale, weigh variables based on importance, and customize the framework based on what matters to your business.
4. Prioritize your conversion tests by using Chris Goward’s PIE framework, which focuses on evaluating three different factors and assigning each a score of 1-10.
This framework will help you understand which pages and elements are the most beneficial and easiest to implement at a glance. It will also assign a score based on potential importance and ease and provide a sum score for reference.
5. Implement Sean Ellis’ FUDWM framework to emphasize mining your data to gain insights, inform your testing, and measure success.
This five-step program will help you understand your visitor’s needs and can be implemented as an automatic feedback loop with insights being collected easily using a tool like Qualaroo.
6. Test using Bryan Eisenberg’s Plan, Measure, Improve framework to dig into the human element, rather than looking at visitor feedback as qualitative data.
This framework will have you ask yourself a series of questions at each step to define test parameters.
7. Use the TIR format to start work on the projects that result in the highest TIR scores first.
Those will have the most lift with the least amount of time and resources. To calculate this score, multiply together the following scores from 1-5: Time: how many days and development hours are necessary for the test to reach maximum impact, 5 being the least amount of time for highest impact. Impact: the amount of revenue or reduced cost that will result from a successful test, where 5 is the best. Resources: the cost of tools, people, etc. needed for this test, where 5 is an indicator of ample and easy access to resources.
8. Invest at least a week between initial research and your live test.
Base conclusions on a sample size of at 250-400 conversions per variation and run tests for at least 4 weeks for statistically significant outcomes. Remember to check your Conversions per day of the week report to make sure you’re not skewing results. For segmented results, you will need thousands of conversions per variation to reach 350+ conversions per variation within each segment.
9. Analyze, learn from, and improve upon your data to make changes.
For example, if more people are watching a video and converting, ask yourself, What does the video cover that the page does not?