📊 Statistics
The maths behind whether a test actually told you anything. Hypothesis design, sample size, power, MDE, and the threats to validity that quietly invalidate results.
📊 Statistics
The maths behind whether a test actually told you anything. Hypothesis design, sample size, power, MDE, and the threats to validity that quietly invalidate results.
🧪 Experimentation
Designing, running, and reading A/B tests without lying to yourself. Foundations, test design, analysis, and the operational layer that turns one-off tests into a programme.
🧠 Conversion psychology
The persuasion frameworks, cognitive biases, and decision-making models behind why CRO tactics work (and why some backfire).
📣 Marketing concepts
Where CRO sits in the broader commercial picture. Offers, funnels, attribution, LTV, and the strategic context that bounds what testing can achieve.
⚙️ Growth engineering
The technical work that compounds across every experiment. Feature flags, instrumentation, personalisation, and the infrastructure that turns ideas into shipped variants.
💷 Business models
How optimisation changes shape depending on what you sell. SaaS, DTC, leadgen, and subscription mechanics - the economics and metrics specific to each, and why a CRO instinct that wins in one model can mislead in another.
These are based on notes I’ve accumulated over years working in CRO and growth - originally fragmented in Obsidian, now polished into something I’d actually share (thanks Claude). I put them together for two reasons.
First, I wanted a clean reference I could send to a client or teammate when a concept came up. Pointing at a link is faster than re-explaining what holdouts are or why peeking inflates the false-positive rate, and shared references are easier to argue against than my own off-the-cuff explanations.
Second, writing things up properly is the fastest way I know to surface gaps in my own thinking. The act of explaining a concept publicly forces a precision that private notes never quite get to.
I’m Alex. I’ve spent 8+ years helping brands like Ford, De Beers, and Hasbro work better - running lean CRO programmes, leading UX, building storefronts from the ground up, and designing content structure for growth. These notes are what fell out of that work, written up properly.