Create flag variations

A flag variation is a reusable group of flag variable values (also called remote feature configuration or remote variables) that let you avoid hard coding values in your code. Instead of waiting for a deployment, you can group them in a flag variation and then update them remotely in the Optimizely Experimentation application.

Optimizely Feature Experimentation dynamically assigns those variable values to your flag based on your flag rules. For information, see Flag variations.

Create flag variations

You must create variables in the flag before assigning them to a variation.

You can create flag variations using the Feature Experimentation user interface (UI) or the Feature Experimentation REST API.

Use the Feature Experimentation UI

Go to Flags > Variations.
Click Add variations.
Configure the variation.
Click Save.

📘
Note
When a new flag is created, the default variations of On and Off are automatically generated.
You cannot delete the Off variation.

Use the Optimizely Feature Experimentation REST API

You can use the Optimizely Feature Experimentation REST API Variation endpoints to manage your variations. See the Create a new Variation endpoint for information.

Brainstorm variation ideas with Opal

You can brainstorm with Opal to automatically create variations.

Go to Flags > Variations.
Click Brainstorm.
Enter the hypothesis you want to test and click send (or press Enter). Opal returns suggestions for variations. See Design an effective hypothesis.
(Optional) When reviewing Opal's variation suggestions, you can:
1. Hover over the variation and click Like response or Dislike response to provide feedback on the suggested variations.
2. Click Add All to add all the suggestions as variations.
3. Click Regenerate for Opal to create variation ideas.
4. Click Exit Chat & Configure to close the Opal chat tab and configure your variations.
Click on a variation to add it. Click Exit Chat & Configure.
Review the new variation and update any suggested configurations, including the suggested defualt variable values.
(Optional) Click Return to chat to brainstorm more variations and repeat steps 4 and 5.
Click Save to add your variations.

Implement flag variations

You deliver flag variations through flag rules to control the end-user's experiences. Flag rules can be a targeted delivery, an A/B test, or a Multi-armed bandit optimization.

In the Optimizely Experimentation application, you see flag variations, which let you group and reuse variable values.

In your code implementation, you can use the returned OptimizelyDecision's variation key to execute code conditionally. But, to take full advantage of variables, you must get the values of the variation's variables directly.

The following code sample shows how you would use flag variables.

📘
Note
This code is indifferent about what type of flag rule the user is in (targeted delivery, A/B test, or Multi-armed bandit), so you can re-use the same code to run different rules.

// Fetch enabled state for the "product_sort" flag.
// Then use flag variations to expose different sorting methods and pagation.

//Define flag variable defaults (if the flag is disabled, you fall back to these values)
sortMethod := "alphabetical"
pageLimit := 10

// Is the flag enabled for the user?
user := optimizely.CreateUserContext(userId, nil)
decision := user.Decide("product_sort", nil)
enabled := decision.Enabled

// In the online Optimizely app, go define flag variations using
// sort_method and page_limit variables.
// For example, variation_1 might sort by popularity and limit results to 15 pages.

if enabled {
	// get flag variable values depending on the variation the user bucketed into
	err1 := decision.Variables.GetValue("sort_method", &sortMethod)
	err2 := decision.Variables.GetValue("page_limit", &pageLimit)
}

// Fetch products using the flag variables, using something like this pseudocode:
products := productProvider.get(sortMethod, pageLimit)

/** 
 * Fetch enabled state for the "product_sort" flag.
 * Then use flag variations to expose different sorting methods and pagation. 
 */

// Define flag variable defaults (if the flag is disabled, you fall back to these values)
var sortMethod = "alphabetical";
var pageLimit = 10;

// Is the flag enabled for the user?
var user = optimizely.CreateUserContext("userId");
var decision = user.Decide("product_sort");
var enabled = decision.Enabled;

/**
 * In the online Optimizely app, go define flag variations using
 * sort_method and page_limit variables. 
 * For example, variation_1 might sort by popularity and limit results to 15 pages.
 */

if (enabled)
{
  // get flag variable values depending on the variation the user bucketed into
  sortMethod = decision.Variables.GetValue<string>("sort_method");

  pageLimit = decision.Variables.GetValue<int>("page_limit");
}


// Fetch products using the flag variables, using something like this pseudocode:
var products = productProvider.Get(sortMethod, pageLimit);

/**
 * Fetch enabled state for the "product_sort" flag.
 * Then use flag variations to expose different sorting methods and pagation.
 */

// Define flag variable defaults (if the flag is disabled, you fall back to these values)
String sortMethod = "alphabetical";
int pageLimit = 10;

// Is the flag enabled for the user?
OptimizelyUserContext user = optimizely.createUserContext("userId");
OptimizelyDecision decision = user.decide("product_sort");
Boolean enabled = decision.getEnabled();

/**
 * In the online Optimizely app, go define flag variations using
 * sort_method and page_limit variables.
 * For example, variation_1 might sort by popularity and limit results to 15 pages.
 */
if (enabled)
{
  // get flag variable values depending on the variation the user bucketed into
  try {
    sortMethod = decision.getVariables().getValue("sort_method", String.class);
    pageLimit = decision.getVariables().getValue("page_limit", Integer.class);

  } catch (JsonParseException e) {
    e.printStackTrace();
  }
}


// Fetch products using the flag variables, using something like this pseudocode:
List products = productProvider.get(sortMethod, pageLimit);

"""
    Fetch enabled state for the "product_sort" flag.
    Then use flag variations to expose different sorting methods and pagination.
"""

# Define flag variable defaults (if the flag is disabled, you fall back to these values)

sort_method = "alphabetical"
page_limit = 10

# Is the flag enabled for the user?
user = optimizely.create_user_context("user123")
decision = user.decide("product_sort")
enabled = decision.enabled

"""
    In the online Optimizely app, go define flag variations using
    sort_method and page_limit variables.
    For example, variation_1 might sort by popularity and limit results to 15 pages.
"""

if enabled:
    # get flag variable values depending on the variation the user bucketed into
    sort_method = decision.variables["sort_method"]
    page_limit = decision.variables["page_limit"]

# Fetch products using the flag variables, using something like this pseudocode:
products = ProductProvider.get(sort_method, page_limit)

# frozen_string_literal: true

# Fetch enabled state for the "product_sort" flag.
# Then use flag variations to expose different sorting methods and pagination.

# Define flag variable defaults (if the flag is disabled, you fall back to these values)
sort_method = 'alphabetical'
page_limit = 10

# Is the flag enabled for the user?
user = optimizely.create_user_context(userId)
decision = user.decide('product_sort')
enabled = decision.enabled

# In the online Optimizely app, go define flag variations using
# sort_method and page_limit variables.
# For example, variation_1 might sort by popularity and limit results to 15 pages.

if enabled
  # get flag variable values depending on the variation the user bucketed into
  sort_method = decision.variables['sort_method']
  page_limit = decision.variables['page_limit']
end

# Fetch products using the flag variables, using something like this pseudocode:
products = productProvider.get(sort_method, page_limit)

/**
 * Fetch enabled state for the "product_sort" flag.
 * Then use flag variations to expose different sorting methods and pagation. 
*/


//Define flag variable defaults (if the flag is disabled, you fall back to these values)
var sort_method = "alphabetical"
var page_limit = 10
 
// Is the flag enabled for the user?
let user = optimizely.createUserContext(userId: userId)
let decision = user.decide(key: "product_sort")
let enabled = decision.enabled 

                                                        
/**
  * In the online Optimizely app, go define flag variations using
  * sort_method and page_limit variables. 
  * For example, variation_1 might sort by popularity and limit results to 15 pages.
 */
  
if enabled {
  // get flag variable values depending on the variation the user bucketed into
  let sortMethod: String? = decision.variables.getValue(jsonPath: "sort_method")
  let pageLimit: Int? = decision.variables.getValue(jsonPath: "page_limit")
	}


// Fetch products using the flag variables, using something like this pseudocode:
let products = productProvider.get(sort_method, page_limit)

// Fetch enabled state for the "product_sort" flag.
// Then use flag variations to expose different sorting methods and pagination. 

// Define flag variable defaults (if the flag is disabled, you fall back to these values)
let sortMethod = 'alphabetical';
let pageLimit = 10;

// Is the flag enabled for the user?
const userId = 'user123'
const user = optimizely.createUserContext(userId);
const decision = user.decide('product_sort');
const enabled = decision.enabled;

//In the online Optimizely app, go define flag variations using
// sort_method and page_limit variables. 
// For example, variation_1 might sort by popularity and limit results to 15 pages.

if (enabled) {
  // get flag variable values depending on the variation the user bucketed into
  sortMethod = decision.variables['sort_method'];
  pageLimit = decision.variables['page_limit'];
}

// Fetch products using the flag variables, using something like this pseudocode:
const products = productProvider.get(sortMethod, pageLimit);

/*
 * Fetch enabled state for the "product_sort" flag.
 * Then use flag variations to expose different sorting methods and pagination.
*/

// Define flag variable defaults (if the flag is disabled, you fall back to these values)
$sortMethod = 'alphabetical';
$pageLimit = 10;

// Is the flag enabled for the user?
$userId = "user123";
$user = $optimizely->createUserContext($userId);
$decision = $user->decide('product_sort');
$enabled = $decision->getEnabled();

/**
 * In the online Optimizely app, go define flag variations using
 * sort_method and page_limit variables.
 * For example, variation_1 might sort by popularity and limit results to 15 pages.
 */

if ($enabled) {
    // get flag variable values depending on the variation the user bucketed into
    $sortMethod = (string) $decision->getVariables()['sort_method'];
    $pageLimit = (int) $decision->getVariables()['page_limit'];
}

// Fetch products using the flag variables, using something like this pseudocode:
$products = $productProvider->get($sortMethod, $pageLimit);

// Fetch enabled state for the "product_sort" flag.
// Then use flag variations to expose different sorting methods and pagination. 

// Define flag variable defaults (if the flag is disabled, you fall back to these values)
let sort_method = 'alphabetical';
let page_limit = 10;
 
// Is the flag enabled for the user?
const decision = useDecision('product_sort');
const enabled = decision.enabled;
                                                        
//In the online Optimizely app, go define flag variations using
// sort_method and page_limit variables. 
// For example, variation_1 might sort by popularity and limit results to 15 pages.
  
if (enabled) {
  // get flag variable values depending on the variation the user bucketed into
  sortMethod = decision.variables['sort_method'];
  pageLimit = decision.variables['page_limit'];
}

// Fetch products using the flag variables, using something like this pseudocode:
const products = productProvider.get(sortMethod, pageLimit);

// Fetch enabled state for the "product_sort" flag.
// Then use flag variations to expose different sorting methods and pagination.

// Define flag variable defaults (if the flag is disabled, you fall back to these values)

var sort_method = "alphabetical";
var page_limit = 10;

// Is the flag enabled for the user?
var user = await flutterSDK.createUserContext(userId: "user123");
var decisionResponse = await user!.decide(sort_method);
var decision = decisionResponse.decision;
var enabled = decision!.enabled;

// In the online Optimizely app, go define flag variations using sort_method and page_limit variables. 
// For example, variation_1 might sort by popularity and limit results to 15 pages.

if (enabled) {
    // get flag variable values depending on the variation the user bucketed into
    var sort_method = decision.variables["sort_method"];
    var page_limit = decision.variables["page_limit"];
  }

// Fetch products using the flag variables, using something like this pseudocode:
var products = ProductProvider.get(sort_method, page_limit);

For more, see the example usage of your SDK:

Create flag variations