- Create and manage Projects and Experiments
- Add, remove, duplicate, and archive Projects or Experiments
- Create separate projects for development and production environments
A project is a way of creating a subsection in your Optimizely account. Each project has its own snippet, set of experiments, and set of collaborators. Projects are designed to help you better organize your experiments across multiple sites (or different sections of large sites). The hierarchy of experiment organization looks like this:
Using Projects, you can group your experiments in ways that make sense for your organization. For example, if you run experiments on multiple domains, you can assign a unique project for each domain. You can also invite other users to collaborate on a project with you without allowing them complete access to your account. Want to see how to manage projects? Watch this short video.
Add a project
To create a new Project, first click the Project list icon ()next to your project name.
Then, click New Project at the bottom of your list of Projects.
To view or manage your Projects, first click the Project list icon () next to your Project name.
You'll see a list of Projects in which you are a collaborator. Click a Project to switch to it.
From this menu, you can also add new projects (if you are an Administrator or Project Owner), or click Manage Projects to see a list of your Projects and their statuses (Active or Archived).
If you're no longer using a project, you can easily archive it. Don't worry, you can always unarchive it later; nothing will be deleted.
Within the Manage Projects page, select the project you would like to archive. Click the Archive button in the right-sidebar and your project will automatically be archived.
Once archived, the status of your Project will change from Active from Archived. To Unarchive, simply click the Unarchive button in the right sidebar.
Create and manage experiments
Each project contains a set of experiments that you run on your site. For each experiment, you'll see the status, number of visitors, number of goals, and dates created and last modified. Want to see how to create and manage experiments? Watch this short video.
Click an experiment to access the Editor, Results page, or Preview Mode for that experiment; or view and edit key components, such as Traffic Allocation, Goals, Audiences, URL Targeting, and the Scheduler.
You can also see its Experiment ID at the bottom of this panel. You'll sometimes use the Experiment ID for troubleshooting. From here, you can also see an experiment's Change History or delete the experiment.
Experiments can have several statuses:
- Running: The experiment is live and running on your site (as long as the Snippet is present).
- Paused: The experiment has once been live, but you are currently not running it on your site. Visitors who were in this experiment while it was live will not continue to see the experiment.
- Draft: The experiment is not yet live on your site.
- Archived: The experiment is not actively in use, and will not be contained in your Snippet.
In the Settings > Privacy menu, you can also disable Paused and Draft experiments from being included in the Optimizely Snippet. This reduces your snippet size and also gives you increased privacy, although it may impede your ability to QA experiments.
To start or pause an experiment from the Home page, click the start/pause buttons.
You may want to create a duplicate experiment when:
- You're using it as the basis of a new experiment.
- You want to direct 100% of traffic to a variation.
- You're copying an experiment from your staging project to your production project (see below).
If you don't want to see an experiment in your list, or you want to reduce your snippet size, click the archive button to archive the experiment.
When you archive an experiment, it isn't deleted; you can always get it back later. Just use the status dropdown to see your archived experiments: change the selection from Active to Archived.
Select the experiment you want to un-archive, then click the archive button again to un-archive it.
Create separate projects for development and production environments
For many organizations, development and production environments are separated so new features can be built and tested without disturbing production traffic. You can use Optimizely Projects to separate experiments in your development environment from those running in your production environment. Creating a separation like this will also allow one or more people in your organization to act as "gatekeepers" for new experiments to be run in production.
Start by creating a new project for your development environment.
Now, grab the snippet for this project, and install that in the head of your development environment pages. Now you and anyone else with access to this project can create experiments on your development pages. Next, create a production project, or simple use an existing project for this purpose. If you wish, you can restrict access to this project to only a certain number of "gatekeepers" who are responsible for ensuring that a test is properly set up before it is deployed in production.
Creating, testing, and deploying experiments
Once you've set up your development and production projects, you can use the following process to create and deploy experiments:
- Create an experiment in your development environment, using development URLs. If your development URLs are the same as those used for your production environment, that's fine as long as your production environment does not use your development Optimizely project snippet.
Edit and test your new experiment to make sure that everything is functioning correctly. For simple experiments, this should be fairly straightforward. For complex experiments, you may wish to take greater care in this step.
Change the targeting of your new experiment to ensure that it runs properly on your production environment URLs.
Duplicate your experiment to your production Optimizely project. Note that this step can only be completed by someone who has access to both your production and development projects.
Now, you can start the experiment in your production project when you're ready!
Learn more about the fundamentals of AB testing.