Conversion Optimization (UX)
Conversion Optimization (UX) tests how to find friction, form hypotheses, and improve completion rates through research and experimentation. Expect funnel math, usability heuristics, A/B testing logic, and prioritization.
Core rules and concepts
- Optimize the whole journey: motivation, clarity, trust, friction, and anxiety. A prettier button cannot fix a confusing offer.
- Diagnose with quantitative and qualitative data: funnel reports, session recordings, form analytics, user tests, and support tickets.
- Write testable hypotheses: because [evidence], changing [element] for [audience] will [effect] measured by [metric].
- Test one meaningful change at a time where possible; define primary metric, guardrail metrics, sample size, and duration before launch.
- Do not stop a test at the first significant result or peek repeatedly; that inflates false positives.
Worked example
A checkout funnel has 100,000 sessions: 50,000 reach product, 20,000 add to cart, 10,000 start checkout, 8,000 purchase. Overall conversion is 8%. The largest absolute leak is product-to-cart (30,000 lost), but the checkout-to-purchase step loses only 2,000. A team might prioritize cart abandonment research if revenue per recovery is high and the fix is feasible. The reasoning: combine drop-off size, revenue impact, and ease-not just the biggest percentage drop.
Traps that cost marks
Ending tests early; changing multiple variables and claiming one caused the lift; ignoring mobile speed and form friction; testing a low-traffic page until it “wins”; and treating correlation from recordings as proof.