Description
Test combinations of changes to several interface elements to learn which mix performs best.
An experiment that tests combinations of changes to several elements of an interface. Create alternatives for elements such as a heading, image and button, allocate visitors to combinations, and compare a predefined outcome such as completed registrations. For example, two headings and three images produce six combinations. Use the results to find an effective combination and investigate how elements work together. Plan for sufficient traffic, since splitting visitors across many combinations can make the experiment slow. It extends the experimental approach of A/B testing to combinations of elements rather than simply comparing complete alternatives.
Attributes
- Discipline
- Experimentation & Causal Inference
- Loss type
- Active
- Stage
- Evaluation, Post-launch
- Data
- Quantitative
- Effort
- Weeks
Videos
Also known as
- MVT
Categories
- Analytics & Behaviour
- Experimentation & Causal Inference