An introduction to Optimize objectives
Objectives are an integral part of creating an experiment that determine how your variants perform and inform the statistical model used to pick leaders.In this article:
This article describes the types of objectives available in Optimize and explains how to configure them. Read on to learn about the difference between Experiment and Additional objectives and the three objectives types available in Optimize.
To configure experiment objectives in Optimize you need:
- An Optimize account and container.
- A website containing the Optimize plugin.
- A new, draft experiment.
- At least one variant.
To create or set an objective, you need:
- Edit access to the Optimize container.
Think of your experiment's objective as "the website functionality you wish to optimize." Common objectives for publishing websites are pageviews and subscriptions, while ecommerce websites tend to use objectives like revenue and conversions.
There are three types of objectives included in Optimize: system, goal, and custom. Adding objectives is done in the configuration card of the Optimize experiment detail view.
Select an objective
In order to run an experiment, you must select an Experiment objective. The Experiment objective is used to determine whether or not a leader has been found for an experiment. In addition to the Experiment objective, Additional objectives can be set in an experiment. Additional objectives allow you to measure your experiment against other metrics, however, they do not inform when a leader is found.
To select an Experiment objective:
- Navigate to the OBJECTIVES tab.
- Select ADD EXPERIMENT OBJECTIVE and add an objective from the drop-down list or create a custom objective.
- (Optional) Add Additional objectives by clicking ADD ADDITIONAL OBJECTIVE.
- (Optional) Add a Description and hypothesis.
- Select SAVE.
Optimize users can use up to three pre-selected objectives per experiment and see data for those objectives in Optimize reports. Optimize 360 customers can select 10 total Additional objectives per experiment. Additionally, Optimize 360 customers can add system and Analytics goal objectives to experiment reports after an experiment has started.
Learn more about the differences between Optimize and Optimize 360.
Optimize supports the following objective types, which are described in more detail below:
- System objectives
- Google Analytics goals
- Custom objectives
System objectives are common objectives that are often used in experiments. Optimize makes system objectives available in all experiments.
System objective types
|Pageviews||System||Pageviews is the total number of pages viewed. Repeated views of a single page are counted.|
|Session duration||System||The length of a session in seconds. A session lasts as long as there is continued activity.|
|Bounces||System||Bounces is the number of single-page visits.|
|Transactions||System||Transactions is the total number of completed purchases on your site.|
|Revenue||System||The total revenue from web ecommerce or in-app transactions. Depending on your implementation, this can include tax and shipping.|
|AdSense impressions||System||An AdSense ad impression is reported whenever an individual ad is displayed on your website. For example, if a page with two ad units is viewed once, we'll display two impressions.|
|AdSense Ads Clicked||System||The number of times AdSense ads were clicked on your site.|
|AdSense revenue||System||The total estimated revenue from AdSense ads.|
Transaction and Revenue are only available when the linked Analytics view is enabled for ecommerce. Learn more about how to enable ecommerce for a view in Analytics.
AdSense system objectives (e.g. Impressions, Ads Clicked, and Revenue) are only available when the linked Analytics view is also linked to AdSense. Learn more about how to manage your AdSense and Analytics integrations.
Goals from linked Analytics view are imported into Optimize for use as an experiment objective. To see detailed Analytics goals in Optimize you need at least Read & Analyze permission to the linked Analytics view. Optimize users without access to the linked Analytics view can still set Analytics goals in an experiment but details such as Goal Name are not exposed.
In addition to system and Google Analytics goals, you can create custom objectives directly in Optimize. Custom objectives allow you to create an experiment objective on the fly. Custom objectives are a good option when an experiment objectives is not as significant as a Google Analytics Goal nor is your experiment objective captured in the system objectives.
Custom objective types
Optimize supports the following custom objective types:
Events objective – Analytics metrics that track user interactions (or “events”) independent of pageviews. For example, video plays or file downloads. An Events objective must include one of the following variables (except URL):
- Event Category
- Event Action
- Event Label
- Event Value
Pageviews objective – The number of pageviews for a particular web page. The URL variable is used when configuring a custom Pageviews objective.
Create a custom objective
To create a custom objective:
- Navigate to the OBJECTIVES tab.
- Select ADD EXPERIMENT OBJECTIVE.
- Choose Create custom from the drop-down list.
- Select an objective type (e.g. Events or Pageviews).
- Create a rule by choosing:
- A variable.
- A match type (e.g. "Equals").
- A value.
- (Optional) Create additional rules by selecting plus ("+").
- Choose a Counting method (e.g. "Once per session.").
- (Optional) Click CHECK RULES in the Rule checker.
- Select SAVE.
- Equals – Every character in your rule, from beginning to end, must be an exact match of the entered value, in order for the condition to evaluate as true.
- Starts with – The starts with match type matches identical characters starting from the beginning of the string up to and including the last character in the string you specify.
- Regex – A regular expression uses special characters to enable wildcard and flexible matching.
- Greater than – The hit value must be greater than the provided value in order for the condition to evaluate as true.
- Less than – The hit value must be less than the provided value in order for the condition to evaluate as true.
Additional rules use the AND operator and will evaluate as true when all conditions are met.
When configuring a custom objective, choose one of the following counting methods:
- Once per session – Optimize counts only one hit with the action per session, even if the action occurred multiple times. This is a good choice if only one unique action brings value to your business. As an example, a Contact Us form generally does not bring additional value if it’s completed multiple times and therefore is recommended to be counted a Once per Session.
- Many per session – Optimize counts every time a hit with the action occurs within a session. For example, all event hits with category equal to Video. This is a good choice if every action, such as a sale, likely brings value to your business.
The counting method you select determines the statistical model used to determine the leader in an experiment.
The values you provide when creating a custom objective must match values in your linked Analytics View. To help you configure a valid custom objective, we display whether or not your configuration matches sessions in your linked Analytics View. The Rule Checker automatically checks the last seven days of data in the linked Analytics view and returns the result of the configured objective. Note that the rule checked doesn’t take into account targeting settings that have already been applied to the experiment.
If the Rule checker returns 0 Sessions, this may indicate that your configuration is not valid. We recommend that you double check the target values for your custom objective before saving your custom objective.
What you see in the Optimize UI is based on your Analytics and Optimize permissions.
You need view access to the linked Analytics view to use the Rule checker or it will appear grayed out. Contact the administrator of the linked Analytics view to verify that your account has view access to the view.
Analytics Permissions: In order to see the details of the Analytics goals and custom objectives you need at least View access to the linked Analytics view.
Optimize Permissions: Users with View access to the Optimize container can see that a custom objective was used in the experiment and the configuration of the objective.
If there are specific ecommerce and/or publisher metrics you’d like to use as objectives in Optimize, select More () in the Optimize app bar then either Send feedback or User forum.
Learn more about new features in the Optimize Release notes.