From the notebook
CRO & Experimentation
Conversion work gets difficult when the numbers look more certain than they are. A statistically significant result can still support a poor decision, and a low-traffic page can still improve without every button colour appearing before a jury.
Here I write about CRO, A/B testing, evidence quality, and the judgment required between the numbers. The aim is to fix obvious friction, test when the setup can teach us something, and say plainly what the available data cannot prove.
What tends to show up here
- Test readiness and interpretation
- CRO with limited traffic
- Evidence before certainty
Notes on CRO & Experimentation
CRO When You Don’t Have Enough Traffic for A/B Testing
A practical low-traffic CRO playbook for marketers: when A/B tests are underpowered, use evidence-weighted decisions, qualitative research, and high-confidence landing page improvements.
A Thought on AI and Conversion Rate Optimization
A practical view from Conversion Boost Copenhagen on AI for conversion rate optimization: reduce hallucinations with structured data, agent workflows, and safeguards.
A/B Testing Is Easy. Interpreting It Isn’t.
A practical A/B test interpretation guide: power analysis, sample size, p-value and confidence interval reading, sequential testing, and SRM checks.