Experimentation for Engineers
summary
Conversion design pairs deliberate hypotheses with A/B tests so engineers measure real impact: experiments with a couple thousand users per variant detect large effects; switching an inaccessible blue form field to an accessible focus state increases signups; swapping oversized images for 250px variants cuts load time and lifts conversion; cleaning components reduces crashes and improves metrics. Mobile experimentation tooling like ABSmartly enables server‑driven UI and faster iterations. Upcoming book Design for Impact and engineering wins around reducing crashes back this approach.
AI-written summary · about
chapters (13)
- 0:00Introduction
- 2:53What is conversion design
- 4:59Linear design process problems
- 7:04Conversion design process
- 10:13Why engineers in ideation
- 13:23A/B testing lowers risk
- 16:53Small decisions compound
- 19:45Effectiveness over speed
- 21:09Experimentation idea buckets
- 22:13Strong experiment hypotheses
- 24:20Objections to experimentation
- 28:38Mobile app experiment examples
- 33:37Summary
related talks
24mEffective Product Experimentation: Using Data-Driven Techniques for Iteration
48mHow to use experimentation in the back-end
52mBuilding an A/B Testing Framework that works for Developers
27mExperimentation in BigTech Validating Hypotheses and Driving Innovation
36mA/B Testing in Production
28mExperimenting your way to a better product
57mHow conversion optimization made me a better interface designer
01h 15mExperimentation at scale: Fuel your application’s growth through A/B testing