Create flag variables

Flag variables let you add and configure more than one variable in a single experiment variation. For example, you can create a feature flag with several variables like color, size, and text. Then, for each variation, you can change one or more of the variable values. This lets you test different combinations of variables and measure their impact on your key metrics.

Using variables offers the following advantages:

• Rapid experiment and delivery deployment – After the code implementing the variables is in place, non-technical collaborators can change the variable values directly in the Feature Experimentation UI.
• Dynamic updates – Changes to flag variable values are automatically reflected in your application (when the datafile is updated, and the user session is re-evaluated), permitting real-time updates to the user experience.
• Consistent variations – All variations must define the same flag variables, ensuring your code can rely on those variables existing in every variation.

Variable types

You can create flag variables of the following types:

• Boolean
• String
• Integer
• Double
• JSON

📘

Note

There is no character limit for a JSON variable value.

Examples of flag variables

For a specific use-case example using an image carousel iOS, see Flag variable example with an image carousel.

General example

For example, you are running experiments and want to improve the number of interactions with a search bar on your website. You could create the following variables and their types:

• Length of the search bar – Integer
• Placeholder copy – String
• Placeholder font color – String
• Location on your page – Integer
• Searchbar visibility – Boolean

You can change the values in these variables without deploying new code. Run different experiments with different variable values to determine the optimal values.

Examples based on role

Depending on your goals and needs, flag variables can be used for various use cases. The following are some examples of different roles, how they could benefit using variables, and some examples of those variables and their type:

Marketers – Improve their conversion rates.

• Copy (headlines, banners, CTAs) – String
• Images – String or JSON
• Colors on landing pages – String
• Layout positions – Integer

Product managers – Improve their revenue and retention.

• Different pricing – Double
• Offers – String
• Features on their product pages – String
• Icons (size, position, labels) – Integer, string, or JSON
• Colors – String
• Messaging – String

Developers – Improve their application user experience and performance.

• Components contributing to different algorithms – Depends on the variable
• Hard capped limits (for example, API rate limits, max objects allowed) – Depends on the variable
• UI components – Depends on the variable

Create flag variable

You can create flag variables in the Feature Experimentation UI or the Feature Experimentation REST API.

Use the Feature Experimentation UI

1. Go to Flags and select a flag.

2. Click Variables and click + Add Variable.

   

3. Select the variable type.

4. Edit the Variable Key and optionally update the Description and Default Value.

   📘

   Note

   You cannot modify the Variable Key after you save the variable.

   

5. Create as many variables as you need for that flag by repeating steps two through four.

6. Click Save.

Use the Feature Experimentation REST API

You can use the Optimizely Feature Experimentation REST API Variable Definitions endpoints to manage your variables. See Create Variable Definition for information.

Brainstorm variable ideas with Optimizely Opal

You can brainstorm with Optimizely Opal to automatically create variables.

📘

Note

If you use Opti ID, administrators can turn off generative AI in the Opti ID Admin Center. See Turn generative AI off across Optimizely applications.

Otherwise, contact Optimizely Support to turn off AI features.

1. Go to Flags and select a flag.

2. Click Variables and click Brainstorm. This opens Opal Chat.

   

3. Enter the hypothesis you want to test and click send (or press Enter). Optimizely Opal returns suggestions for feature variables when running your experiment. See Design an effective hypothesis.

   

4. Click on a variable suggestion.

5. Review and update the variable details pre-filled by Optimizely Opal.

   

   📘

   Note

   You cannot modify the Variable Key after you save the variable.

6. Click Save to create the variable.

7. (Optional) Return to Opal Chat and click on a variation and repeat steps five and six to create additional variables.

Implement variables

After you have created your variables, the next step is to create flag variations. Variations let you avoid hard-coding variables in your application and let you group and reuse variable values using flag rules.