This article focuses on the Manager's main features. For in-depth information on the different features in the Manager, refer to the linked articles at the end of each section below.
The Evolv AI Manager (the Manager) is a web-based platform where you can manage, configure, and analyze the performance of your optimization projects. It’s also from the Manager that you can connect the Evolv AI product ecosystem to your digital experiences.
Within the Manager, you have control over your optimization projects, with features ranging from user access management and governance to configuring audience segments for experiments and deploying integrations with third-party tools. It's also where you can closely monitor your optimization project's performance, analyze results, and derive insights.
Accounts in the Evolv AI Manager are attributed to distinct Evolv AI customers. Each account can have many sub organizations, typically attributed to different business units within an account, such as “Internet” and “Mobile” services.
Each account organization has its own environments that can track audiences and user behaviors independently of other organizations. User accounts are attributed at the organization level as well.
Within an organization are the individual optimization projects.
A project is a means of organizing and managing optimization ideas and analyzing their performance for a specific audience and metric.
Projects typically differ by the audience and/or metric they are designed to optimize.
There are 4 key aspects that define a project:
- Experimentation Type – The Experimentation Type is the way in which a project is expected to be run, using either an A/B testing or Continuous Optimization methodology.
- Project Audience – The Project Audience is the primary filter for including visitors in a project.
- Optimization Metric – The Optimization Metric is the parameter used to measure the performance of the project audience.
- Variables & Variants – The ideas you have to optimize your digital experience can be expressed in two parts:
- Variables – A variable represents an aspect of the digital experience that can vary, or change. Functionally, variables are like A/B tests, they include a control variant that is used to compare the performance of one or more non-control variants.
- Variants – A variant represents a distinct version of some aspect of the digital experience that you want to test.
Events are actions or outcomes that can occur during a visitor's session.
On the Events page, you can create and configure events to record and measure during a visitor’s session. These recorded events form the basis for analyzing visitor behavior and making optimization decisions.
Optimization Metric and the Conversion Event
In a project, the Optimization Metric is the specific behavior you want to optimize in your digital experience. It is defined by a conversion event and optionally, a conversion value. In a Continuous Optimization project, the system uses the conversion event to make decisions about what to continue testing or not.
Audiences represent different segments of visitors in your digital experience.
On the Audiences page, you can create audiences that are defined by unique targeting criteria. Defining audiences allows Evolv AI to track their behavior and measure the rate at which they fire different events. In a project, the AI uses this data to make predictions about which optimization ideas will have greater impact on an audience.
Audiences can also be used to test personalization ideas within a project.
Integrations allow the Manager to interact with other applications, enabling data exchange and enhancing insights and decision-making. You can ingest data from or share data with other applications, including web browsers and third-party analytics tools (e.g., Google, Adobe).
The Manager offers support for both public integrations (such as Google and Adobe Analytics) and custom integrations (using NOM, GitHub, Gist) using APIs to connect with any tool capable of ingesting data.
A custom integration can be used to run code on any page the Evolv AI snippet is loaded so you can run code outside of projects, such as firing events.
- Integrating with Google Analytics 4 (GA4)
- Integrating with Adobe Analytics
- Creating a custom integration
An environment signifies a specific configuration or context in which software applications or systems operate. Environments are instrumental in managing different stages of software development, testing, and deployment. They ensure that software functions consistently and accurately across various scenarios.
In the Manager, environments act as unique connection points, establishing a distinct link between Evolv AI and your digital experience. Optimization projects are published to environments and data is collected and processed at the environment level.
Learn more: About Environments
The Schema in your account organization tells the system how to process data about a user that it receives from different sources.
The schema applies to all environments in your Evolv AI account organization, and consequently, all projects share the same schema configuration.
The data defined in the Schema can be used in two ways:
- Targeting visitors based on their context attributes eg. cart_age, return_status
- Optimizing numeric attributes eg. order_value, video_minutes_watched
- About the Schema
- Optimizing the value of context attributes
- Targeting audiences with context attributes
The Users section of the Manager allows the administrator of an Evolv AI account to manage user access and assign roles. There are four user roles within the Manager:
Access and permission roles are controlled at the Organization level within an Account, where only Admins can invite users via email.
Users are able to access the Manager through Google SSO, just one of many ways Evolv AI helps users ensure the security of their data.
Learn more: Inviting a User to an Organization