Analytics without false confidence
At WordCamp, I spoke about conversion optimization and online analytics. The talk came from a recurring frustration: many teams discuss A/B testing and conversion data without enough statistical judgment to know whether a result is meaningful.
What the talk covered
- Why a measured difference is not automatically a reliable difference.
- How sample size and uncertainty change what you can conclude from an A/B test.
- Why optimization work needs both measurement discipline and practical business context.
The operating point
Analytics should make decisions sharper. If the data cannot support the conclusion, the dashboard is only decoration.