A/B Testing
A/B testing is when two different design options are tried for a website. Each option is served to half of visitors. That visitor’s engagement is then measured based on the selected criteria (such as registrations, subscription purchases, etc).
An example is using two different labels for your Call To Action button on your landing page to see which label drives more registrations.
What To A/B Test
Ideas you have with a medium amount of confidence should be A/B tested. If your confidence is low in an idea, gather more evidence and postpone testing it. If your confidence is very high in an idea, then just implement the idea.
Prioritizing testing
GoodUI prioritizes based on three criteria:
- Highest Expected Effect + Highest Confidence
- Lowest Effort
- Shortest Testing Duration
In other words: aim for maximum juice-to-squeeze.
References
- Should You A/B Test Your Idea, Or Not? (GoodUI)
- What To A/B Test Next (GoodUI)