1. Help Center
  2. Getting Started

Full Stack Experiments on Convert

THIS ARTICLE WILL HELP YOU:


Creating a full-stack project on Convert involves integrating Convert's A/B testing capabilities into both the frontend and backend components of your application. This approach allows for comprehensive experimentation, enabling you to test and optimize the entire user experience, from the user interface to the underlying business logic and database interactions. Below is an updated step-by-step guide, now including details on setting up and launching experiments using the Node.js SDK.

1. Understanding Convert and Full-Stack Projects

Convert specializes in A/B testing, offering tools to create experiments that test different versions of web pages or applications to improve engagement, conversion rates, and performance. A full-stack project involves both client-side (frontend) and server-side (backend) components, including databases. Integrating Convert to these layers enables holistic experimentation across your application.

2. Setting Up Your Account and Project

  • Sign Up for Convert:
  • Begin by signing up for a Convert account at Convert's website.
  • Create a Full Stack Project:
  • In your Convert account, create a new project specifically for your full-stack application. This project will house your experiments.

    Projects List - Convert

3. Creating a Full Stack Experiment

Initiate a Full Stack Experiment: Within your newly created Full Stack project, proceed to create a Full Stack Experiment. This setup allows you to define the parameters and goals of your experiment, tailored to both frontend and backend aspects of your application.

Experiences - Full Stack Demo - Convert

4. Integrating Convert with the Frontend Using the Node JS SDK

Install the Convert SDK:

To conduct client-side experiments, the Convert Node JS Full Stack SDK must be installed in your frontend. This can be done by using npm or by incorporating a JavaScript snippet, as specified in the SDK documentation. For experiments to be launched and goals to be tracked, modifications to your website's code are necessary. These changes enable the initiation of experiments and the tracking of user interactions directly through the SDK.

Set Up and Launch Frontend Experiments:

The configuration of your frontend experiments is performed within Convert's dashboard, where you can define variations and set specific goals, such as improving click-through rates or conversion metrics. However, to activate these experiments and track their performance, you must modify your website's code to integrate with the Full Stack Node JS SDK. This integration allows for the dynamic adjustment of your site's content and functionality for A/B testing, ensuring that you can directly manage and observe the impact of your experiments on user behavior and goal achievement.

5. Integrating Convert with the Backend Using Node JS SDK

  • Full Stack Node.js SDK Integration:
  • For backend experiments, integrate Convert's Full Stack Node.js SDK. This enables you to apply experiment variations to server-side logic, database queries, or other backend functionalities.
  • Launch Experiments Server-Side or Client-Side:
  • With the Node.js SDK, you can launch experiments either server-side or client-side. However, a Node.js environment is required for the SDK to function and for launching your experiments.
  • Experimentation Logic:
  • Implement logic in your backend to manage the different experiment variations, which may involve serving varied content, applying different business logic, or altering database interactions based on the user's experiment variation.
  • Tracking and Goals:
  • Ensure your backend can track user interactions and report these back to Convert, involving server-side tracking of actions, conversions, or other relevant metrics.

6. Running Full-Stack Experiments

  • Launch Experiments: With the frontend and backend integrations complete, including the Full Stack Node.js SDK setup, you can launch full-stack experiments. Start the experiment in Convert to serve variations to users as configured.
  • Analyze Results: Monitor each variation's performance using Convert's analytics tools, assessing impacts on both frontend and backend metrics.

7. Best Practices

  • Incremental Changes: Start with minor changes to minimize risks and simplify performance impact analysis.
  • Comprehensive Testing: Thoroughly test all variations to avoid negatively impacting user experience or system functionality.
  • Continuous Learning: Use insights from each experiment to inform future development and testing strategies.

Conclusion

By integrating Convert into a full-stack project, including the use of the Node.js SDK for backend experiments, you can optimize every application aspect. This comprehensive approach to A/B testing enhances user experience, conversion rates, and overall system performance.