It felt faster and more creative, but over time we realized we couldn’t clearly explain why certain changes worked. That’s when we shifted toward a more performance-based mindset. I remember reading a detailed breakdown about how some teams structure CRO around measurable impact rather than design trends — this one in particular stood out
https://wingetui.com/how-conversionr...nce-based-cro/ . What clicked for me was the focus on structured hypotheses and continuous experimentation instead of random updates. Once we started tracking user behavior more closely and running controlled experiments, improvements became more predictable and scalable. It takes discipline, but it removes a lot of guesswork.