The messy science of conversion rate optimization

The messy science of CRO looks at how experimentation actually works inside real organisations not laboratories. While CRO borrows heavily from scientific methods, applying “gold standard” rigour everywhere can slow learning, reduce impact, and create false confidence.

In my talk, I argue for proportional rigour: adapting how rigorous we are based on context rather than chasing scientific purity. Using examples from CRO, psychology, and medical research, I introduce five practical lenses (stakes, learning, signal vs noise, team capability, and decision-making confidence) that help teams choose the right level of rigour for the decision they’re trying to make.

The goal isn’t to run perfect experiments every time, but to build practical understanding, compound learning over time, and make better decisions under uncertainty in the messy reality of product and UX work.

‍