This article covers the following: |
Overview
VWO Fullstack has evolved into VWO Feature Management & Experimentation (FME), offering several new capabilities. FME is a one-stop solution for managing features, executing progressive rollouts and automating guarded releases, running A/B tests, and delivering personalized experiences.
It helps understand feature impact, accelerate shipping velocity, and ensure deployment safety—an approach trusted by the best tech-driven companies.
FME supports multiple SDKs and frameworks across digital platforms, such as web, mobile, tablet, and kiosks, catering to different development environments. The complete list and their specific implementations are here.
What's New in VWO FME (for existing VWO Fullstack users)?
The following table includes a list of all capabilities that VWO FME offers:
New updates | How does it help? |
Unified feature management | Control your feature releases using advanced feature flags tailored by environment or user segment. Under the same flag, run rollouts, testing, or personalization rules. |
Dynamic configurations | Instantly change app behaviors with JSON-supported feature flags, eliminating the need for frequent app updates. |
Advanced segmentation | Target user groups by sending custom attributes or events from your codebase or third-party tools. You can also include the user agent to target users based on browser details such as location, browser name, OS, and more. |
Priority management | Prioritize experiments, rollouts, or personalization campaigns to avoid conflicts and streamline deployments. |
Rollout & rollback automation | Perform progressive rollouts, automate kill switches based on defined metrics or timelines, and evaluate impact (feature on vs. off). |
Impact analysis | Measure the true impact of your feature rollouts through integrated analysis of key metrics, both upstream(e.g., ad clicks) and downstream(e.g., revenue). |
Tech debt management | Automatically detect and alert you about unused flags, helping keep your codebase organized and clutter-free. |
User behavior analysis | Combine with VWO Insights to watch session replays, deeply examining user actions during feature rollouts. |
Zero latency decision-making | Instant decision-making directly on your backend servers, mobile app clients, or front-end browsers, eliminating delays. |
Scalable architecture | Built to support continuous CI/CD pipelines, offering enhanced scalability with over 12 SDKs compliant with the OpenFeature standard. |
Efficient variation evaluation | Achieve faster variation evaluations, significantly reducing API calls and improving overall performance. |
Integration | Connect to your data warehouse, CDP, CMS, and more to analyze the impact of your features and experiments as you want. |
Feature debugger | Use a powerful debugger to test feature flags, evaluate rules, and conduct dry runs before full deployment. |
Rest APIs | Manage feature flags with powerful REST APIs. Create, update, and retrieve flag configurations (e.g. variations) easily—improving control and seamless integration into your workflow. |
Key Concepts in VWO FME
This infographic illustrates the end-to-end workflow of using VWO FME—from flag creation to analyzing the impact through detailed reports.
This section further explains each key concept in detail.
Feature Flags
Feature flags enable you to control when and how new features are shown to users, independent of your deployment cycles.
Each feature flag includes a default or control variable. You can define multiple variables per flag.
Variables can be set to different types—Text, Number, Boolean, or JSON (especially useful for dynamic, complex configurations).
This flexibility allows you to manage various scenarios, from simple feature toggles to intricate logic adjustments using VWO.
Variables
Variables help you define customizable elements within your feature flags—like UI components, algorithms, AI models, or specific logic paths—without redeploying your application.
You can modify these components instantly through the VWO platform by setting them as variables. This helps you respond rapidly to user feedback, adjust configurations, and iterate quickly on features.
Variations
Variations allow you to assign values to your defined variables. Different variable values show different variations of a feature.
By comparing how users engage with each variation, you can clearly identify which option performs best. This will help your team make data-informed decisions and optimize feature performance.
Metrics
Metrics let you measure the real-world impact and effectiveness of your feature releases or experiments, helping you see what’s working.
They’re calculated directly from events triggered by user actions in your app or website. For example, when a user views a page or clicks a button, each action is recorded as an event.
Metrics aggregate these individual events into meaningful data—like total page views, button click counts, or conversion rates—giving your team clear insights into feature performance.
Rules
Rules let you control precisely who experiences a feature, participates in an experiment, or receives personalized content.
You can target users based on attributes (e.g., location), behaviors (actions users have or haven’t taken), or even roll out features to a defined percentage of your audience within a specified time frame.
- Rollout Rule: Gradually expose new features to controlled groups of users, minimizing risk while allowing your team to monitor performance carefully before a broader release.
- Testing Rule: Easily conduct A/B tests, delivering different feature variations to your users. Quickly determine the variation that achieves your desired outcomes by comparing performance metrics.
- Personalize Rule: Give customized experiences tailored specifically for certain user segments, ensuring your content or features resonate more deeply and drive stronger user engagement.
Step-by-Step Guide to Getting Started
Before starting these steps, make sure you have an active VWO account. If not, you can sign up for a 30-day free trial.
Follow this guide to set up VWO’s FME efficiently, empowering your team to manage features confidently, run tests, personalize experiences, and streamline rollouts.
Step 1: Set up your feature flag
- Create Feature Flag: On the VWO dashboard, navigate to Feature Management and create your new feature flag.
- Define Variables: Decide what aspects of your feature you want dynamic (e.g., UI text, button colors, algorithms) and specify their data types (Boolean, text, number, JSON).
- Create Variations: Set up multiple variations with different variable values to effectively test or personalize your feature.
Step 2: Define metrics
Identify and configure metrics (e.g., clicks, conversions, response time, error rates) to measure your feature's performance and success.
Step 3: Establish rules
Choose and create rules according to your rollout strategy:
- Rollout Rule: For progressive rollout to a limited audience (e.g, 10% of users)
- Testing Rule: To run controlled experiments with variations (e.g, test pricing variations)
- Personalize Rule: Customize features specifically tailored for user cohorts (e.g, personalize UX for premium users)
Step 4: Integrate SDK and implement code
This infographic illustrates how VWO dynamically evaluates feature flags at runtime to deliver the appropriate user experience based on flag status.
To integrate the SDK and implement the code:
- Integrate the relevant VWO SDK into your codebase. Here is the complete list.
- Initialize the SDK using the init() function near your app’s existing setup code.
- Set up a polling interval or implement a webhook to ensure the SDK always receives the latest changes.
- Before your new feature’s code, call the getFlag() function, passing user information through the userContext object. This checks if the feature is active for this specific user.
- If the feature flag is active for the user, execute the new feature’s code block. Use getVariable() to fetch different configurations tied to the feature. The variable values depend on the feature flag rules and the variation assigned to the user based on those rules.
- If the feature flag isn’t active, execute the original code block without the new feature.
For more information about the implementation, refer to our SDK documentation.
Practical Use Cases
Example 1: AI model testing
Hypothesis | Adjusting AI chatbot temperature settings improves user engagement and the naturalness of interactions. |
Feature Flag | AI_Response_Temperature_Test |
Feature Variables |
|
Impact Metrics |
|
Rules |
Testing Rule: 3 variations
|
Example 2: Search algorithm optimization
Hypothesis | Search algorithm using recent user activity boosts click-through rates and enhances user experience. |
Feature Flag | Activity_Based_Search_Algorithm |
Feature Variables |
|
Impact Metrics |
|
Rules |
Testing Rule: 3 variations
Personalize Rule: After testing, personalize algorithm type for segments based on activity levels |
Example 3: Improve user onboarding flow
Hypothesis | Simpler onboarding with personalized tips reduces drop-offs and improves retention. |
Feature Flag | Simplified_Onboarding_Flow |
Feature Variables |
|
Impact Metrics |
|
Rules |
|
Need more help?
For further assistance or more information, contact VWO Support.