The web is not static, and neither are your competitors. Your web analytics team should always be looking to create hypothesis to optimize your website. Optimization is about finding areas of your site that can be improved on a systematic basis, and testing the validity of those hypothesis.
Web Analytics Hypothesis
Web analytics management needs to provide your team with the necessary software and tools to conduct proper analysis and create hypothesis. This can be mouse tracking software, deeper dive software such as Omniture Discover, survey software such as Forsee, and social measuring tools. These tools need to be used in conjunction with your web analytics platform. Providing your team with many tools to analyze pages will help them validate their hypothesis and create use cases.
Web Analytics Validation
This one is very important, as your web analytics manager should validate if the hypothesis are correct statistically and organizationally. Web managers need to validate the quality of the hypothesis by statistical means, making sure that significant data was collected. Web analytics managers also need to validate that the hypothesis aligns with organizational goals.
For example, analyst came up with the hypothesis that the Green Widget page needs to have a form on right rather than a form on left. This was done by analyzing the mouse tracking software. Web analytics manager needs to review that the eye tracking software had significant visits to create this hypothesis. More importantly, the manager needs to ask product managers if the organization is pushing to sell more Green Widgets. Any misalignment in the validation phase causes further testing, optimization and measurement for a possibly discontinued product.
Website Testing, Optimization, Measurement
Validated hypothesis need to be tested. Proper A/B or Multivariate testing techniques need to be conducted to make sure that the hypothesis will provide significant lift for your organization. Once tests are conclusive, most optimization software allows you to leave the winning version on your site until your website team has updated the code with the changes. During this process, your web analytics team should be measuring the results of the update and documenting the data for sharing.
Sharing Results With Executives, and Product Managers
Once your team has been through the optimization cycle several times, you are ready to share results. This is a very important step in driving a data driven organizational culture. Product managers and executives who see successful improvement in website performance become advocates of the structure and process. New hypothesis for tests are developed and validated. Your web analytics is now driving significant value to your organization.
Conclusion
With everything in data there are caveats. The most important caveat to the above process is your web analytics tags have to be deployed correctly and data has to be validated. Incorrect sourcing of data leads to distrust and breaking of the web optimization process.
Related Posts:
3 ways to find something to A/B Test
How to become a great web analyst
When to switch your web analytics platform