Skip to content

Count page views and project clicks with PostHog (EU servers)? It loads only if you choose Allow. Privacy

A/B Testing live preview
All projects

A/B Testing

Practise reading an A/B split before paying for an experimentation tool.

Enterprise A/B testing tools are priced for teams, and a solo builder rarely gets to practise reading a split before running one.

Tech Domain · Book Dev|Published|Evidence: Prototype|Next.js, TypeScript

A public prototype demonstrates the interaction; scope and data are intentionally limited.

Case Study

Before

Reading an A/B result well means understanding weights, exposures and conversion rates, and enterprise tools are too expensive just to practise with.

After

Split-test bench on the Dashboard shell: change the split, simulate 1 or 50 visits, add conversions and watch the per-variant rate bars move; ties show no leader, and state is saved in this browser's localStorage.

How it's solved

Built a local split-test bench with three seeded variants (Control, Signal, Proof), a traffic-weight slider per variant with an allocation check, weighted-random visit simulation, a manual Add conversion button, and stat tiles for exposures, conversions, overall rate and current lead.

Trade-offs

Simulation only: no real visitors, traffic splitting on live pages, or statistical significance test, and the UI says so.

Related Articles

Further reading on related topics from the blog

Related Projects

More in Tech Domain · Book Dev