<aside> 💡 Why ? : This doc list best practices to implement experiment in a company.

</aside>

Table of contents

As Ron Kohavi states in his book, It’s critical to understand that

leadership cannot just provide the organization with experimentation platform and tools. They must provide the right incentives, processes and empowerment for the organization to make data driven decisions.

1 - Have the right company data culture

Prior to any AB testing plan, the company should be really crystal clear and honest about it’s current data maturity status. Everyone should have visibility on the metrics and be aware of their importance. Eg; A software engineer that doesn’t know that he is working to improve a metric on conversion will not have the intention to implement tracking and test features. He will be frustrated to drop a feature because a data analyst said so.

That’s why, leadership should be really data driven, and so the company culture. Main elements to validate are :

• The Hippos (highest paid person opinion) is not a standard for decision making, but it’s data
• High level company goals and KPIs are shared and goals are clear
• Each individual know what metrics he is supposed to work toward (define Team MECE tree)
• Empower teams on improving metrics and test innovative ideas. With culture of testing and failing fast
• Have agile organization to create healthy quick feedback look to iterate on experiments

Few actions you can take to grow this culture :

1. Communicate metrics on common channels in a comprehensive manner (slack)
2. Allow user to get notifications about their metrics on their mail
3. present to tech teams results of analysis