Goals Overview
Overview
A goal is what you're measuring in an experiment - a sale, a click, a pageview, a scroll depth, or something custom you define. Every experiment can have as many goals as you need.
Local vs. Global goals
This is the most important distinction to understand about goals in Explore.
Local goals apply only to the experiment they're created in. There's currently one: Track Clicks, created directly inside the Visual Editor.
Creating a Track Clicks goal requires the browser extension, since it's built inside the Visual Editor.
Global goals are created once and can be reused across any experiment. There are four: On Click, Pageview, Scroll, and Custom. These are created from a form in the Goals menu and don't require the extension.
Track Clicks and On Click do the same thing - track clicks on an element - but Track Clicks is Local (tied to one experiment, created in the Visual Editor) while On Click is Global (created via Advanced → Goals, reusable everywhere, and can track the same element across multiple pages).
The 3 default goals
Every experiment comes with these three goals already added:
-
Bounce Rate - tracks how many visitors left the page without interacting with it in any way. No integration required. See Bounce Rate Goal.
-
Engagement - tracks how many visitors viewed more than 3 pages and stayed more than 180 seconds. No integration required. See Engagement Goal.
-
Sale - tracks purchases made by visitors who saw the experiment, along with average order value and revenue per visitor. Requires integration. See Sale Goal.
Creating a Global goal
-
From the left menu, click Advanced, then Goals.
-
Press New Goal.
-
Name it, add a description, and decide whether to allow multiple conversions per visitor session (see below).
-
Select the Goal type: On Click, Pageview, Scroll, or Custom.
-
Fill in the type-specific details (covered in each goal's own article).
-
Press Save.
Finding an element's selector (for On Click goals)
If you're not comfortable with code, here's how to find any element's selector using your browser's Developer Tools (Ctrl+Shift+I, or via the 3-dot menu → More Tools → Developer Tools):
-
Open the Elements tab and press the Element Inspector button.
-
Click the element on your page you want to track - it highlights in the Elements panel.
-
Depending on which selector type you want:
-
HTML Tag Selector - the value before
class(e.g.button). Tracks all elements sharing that tag. -
HTML Element ID Selector - the value after
id(e.g.8a72a2f). Tracks only that exact element, but not every element has one. -
HTML Class Selector - the value after
class(e.g.omni-book-a-call-trigger__button). Tracks all elements sharing that class - but an element can have multiple classes, and you can only use one per goal. -
CSS Selector - right-click the highlighted element, choose Copy → Copy selector. This is the most reliable option and the one we recommend for On Click goals.
-
-
Paste the selector into the goal screen and press Save.
Using a goal in an experiment
Once created, add a goal to an experiment from its Traffic Allocation & Goals section - either select an existing goal, or create a new one directly from that screen.
Creating a goal doesn't automatically add it anywhere - you always need to manually attach it to each experiment where you want it tracked.
Goal Timeout
Sets a time limit, in days, for registering conversions after a visitor sees the experiment. Found in Traffic Allocation & Goals, at the bottom.
The default is 3 days - meaning only visitors who convert within 3 days of seeing the experiment are counted. This applies to every goal on the experiment, not per-goal.
Multiple conversions per session
When creating a goal, you can enable allow multiple conversions per visitor session. With this on, the same visitor can trigger the goal more than once in the same session, and each will be counted.
Not every goal type supports this - Bounce Rate and Engagement don't, since they represent one-time states rather than repeatable actions.
Verifying a goal is working
Before trusting your data, it's worth confirming a goal actually fires correctly. See QA a Goal for the full verification process.