Traffic quality can make almost every other metric look better or worse, so I wanted to open another thread on how people evaluate it in practice.

The specific challenge I am seeing is inconsistent session quality.

I am trying to build a process that is objective rather than emotional, especially when a traffic source initially looks promising.

• Looking at conversion behavior, not just clicks
• Trying to separate weak traffic from fraudulent traffic
• Open to both manual checks and automated filters