Develop a continuous, iterative process to increase sales online.
Conduct as much qualitative and quantitative research as possible,before every optimization attempt.
This includes qualitative research like surveys and user testing, as well as quantitative data from your analysis setup.
Build hypotheses to give your testing program direction and allow for increased learning, regardless of the results.
To create a hypotheses undertake the following steps:
- Use Google Analytics to find problem pages based on where people are abandoning the website.
- Use heat maps and session recordings found in a tool like Hotjar, to help you ascertain where things might be going wrong on the page.
- If more information is required to understand the problem, consider running usability testing on the page.
- Make an educated guess as to one or more ways these problems could be addressed. These are your hypothesis.
Where only simple changes are required like changing some text, this can be done using an A/B testing. For more complicated changes, you may wish to build a prototype in a tool like Balsamiq, Sketch, or Figma.
For simple changes, you can test using an A/B testing tool like Google Optimize. These tests will provide highly accurate quantitative results, but do require significant traffic to gain statistically relevant numbers. You can test your prototype for more complicated changes through usability testing or using a platform like Maze.
Every test should provide deeper insights and lead to new hypotheses.