Introduction

Suggested Elements uses AI to automatically uncover and label elements, reducing manual effort and providing insight into the usage of key website components. This feature streamlines the process of understanding user interactions with your website.

Getting Started with Suggested Elements

To view Suggested Elements, go to Settings > Elements, and click on View Suggestions. This provides a list of Element names.

You can then click through to review the Element in Data Studio.

Each suggestion includes:

• Name: Recommended label for the element.
• Selector: Web or mobile CSS identifier for the element on your webpage.
• Description: A sentence describing the purpose of the element.

Within Data Studio, you can Accept or Reject the suggestion. The suggestion can be modified, or can be accepted as-is. After a suggestion is saved, it can be used for analysis just like user configured Elements.

FAQs

How does AI generate Suggested Elements?

Suggested Elements are created by analyzing Fullstory’s comprehensive session data. Events from sessions with unlabelled data are proactively analyzed and classified by AI to produce suggestions.

How often are suggestions produced?

Suggested Elements are produced multiple times a day, and a max of 10 Elements at a time will be provided. Suggestions can only be manually generated if all pending suggestions are reviewed.

I’ve already configured a similar Element as an incoming suggestion, what should I do?

Suggestions are only produced on components that do not match a configured Element. If a configured Element with the same purpose exists, it is likely that the existing Element selector is too specific.

Common solutions include:

• Adding the suggested selector to the existing Element, and modifying the name or description as is necessary. The suggestion can then be rejected.
• Saving both Elements, and modifying their names to distinguish their purposes.

Do suggestions expire?

Suggestions can be reviewed as long as there are still events that are coming in, but will be removed from the list if an event with a matching selector has not occurred in over a week.

Processing data for Suggested Elements

Event data is processed by Google Cloud models in our existing GCP infrastructure. The data processed consists of the events in sessions with unlabelled data. Producing suggested elements does not involve or necessitate training any models on customer data. Suggested elements is one of many AI features Fullstory is experimenting with that will empower teams to work smarter. Browse other AI features and opt in or out of them for your organization within the Settings > GenAI Features area of Fullstory.