This section allows you to change the following parts of an experiment:
- The Experiment's Name
- The Type of Experiment
- The actual changes using the visual editor, or the URLs representing the variations of your experiment.
For the experiment’s name you can choose any name you desire, that helps you identify the experiment.
Type of Experiment
You can choose from 5 options that Convert allows you to perform, which are the following:
- A/B Experiment: In this experiment you create the changes or different versions of the pages that you want to test in the experiment. In this experiment the changes are hosted and served from Convert servers. We then measure the goal conversions that you have setup on your experiment.
- Split URL Experiment: In this type of experiment you supply the URLs of the different page versions that you would like to test. We split the traffic between the different version and measure the goal conversions that you have set up on the experiment.
- Multivariate Experiment: In this experiment you create different versions for sections of your page (such as a footers, or headers). Convert mixes all this versions of the different sections of your page and serves them to the visitors. At the end, Convert tells you what are combinations of the page sections have a better conversion rate.
- A/A Experiment
The traffic allocation configuration allows you select the percentage of the visitors of your site, that should be put in the experiment.
The traffic allocation percentage will be served in equal proportion the variations that are included in your experiment and measure its corresponding conversions on the goals that you have included in the experiment.
The rest of the visitors won't be put on the experiment.
This might be an attractive option for websites with very large traffic in which it is preferred to not put all its visitors on the experiment.