Creating your first experiment
Sidecar allows you to create and run A/B experiments easily without having to write code or push code to production. Here we'll see how you can create one such experiment and get results
This guide assumes you have followed the previous steps of installing side-car, creating a statsig account and setting up the API Keys correctly. Check out Setup for those instructions.
Step 1: Navigate to the web page
Navigate to the web page you want to experiment on.
Step 2: New experiment
Hit the New Experiment button and fill out the details. This will create a local experiment which hasn't been published yet. This allows you to configure all the details, verify that everything works and then you can publish it.
Once it's created your Sidecar panel would look like this. Notice that the Experiment is now in "Draft" state.
Step 3 (Optional): Add url filter
You have the option to select what pages this experiment is valid by choosing a regex. This regex will be matched against page urls, and the experiment will only be active for those urls that match.
For example, (http|https):\/\/www.statsig.com\/pricing matches pages http://www.statsig.com/pricing or https://www.statsig.com/pricing, and will activate this experiment on those pages.
Step 4: Add actions
Click the Add Action button and you'll see a list of actions you can perform with this experiment. Let's try one of them here.
Go ahead and choose Change content of an element. This will set you up to run an A/B test changing the content of an HTML element - like Headlines, descriptions, CTA, etc.
Step 5: Select an element
In order to run a content change experiment, you will need two things: 1. the element that you want to test with, 2. the content you want to change.
Click on the yellow Target element path text-box. This will activate an element selector mode.
Now as you move your mouse over your web page you'll see a red selection rectangle. Choose the element you want by clicking on it. In this example, we're choosing the main Headline.
Sidecar will now reflect the path of the element that was selected.
Step 6: Update content
Now, you can choose the two different text content you want to A/B test. In the Control content text box, add your control text ("Build Better Products") and in the Test content text box, add your test variant ("Experiment Like a Pro").
You can validate these changes in realtime by clicking on the ▶ button above the text box for each variant. This will immediately change the element's content so you can visually inspect how things look before publishing.
Step 7: Add more actions
Feel free to add more actions within the same experiment and play with the capabilities of the tool