Experiment Overview

Experiments in Performance Now! let you dynamically test templates, URL parameters, or CSS changes—enabling true optimization across landing pages, user journeys, and even visual appearance.

Experiment Fields

• Name:
  The display label for this experiment.

• Status:
  Indicates whether this experiment is live. Only active experiments serve variants or log results.

• Key:
  Public identifier used in all reporting and tracking.

• Type:

  • Template: The default mode for most experiments. Use this to run standard split tests, swap content, or adjust layouts.
  • Parameter: Lets you rewrite an incoming URL parameter dynamically. For example, to test multiple color values (?color=red), set key to color and define all values to test as outcomes.
  • CSS: Use to dynamically inject CSS rules as outcomes—enabling visual/UI tests without dev deployment.

Outcomes

Each experiment can have multiple outcomes (variants) associated with it. For parameter and CSS types, outcomes control what value or rule gets applied when a user lands on the page.

Metadata & Grouping

• Metadata:
  Each outcome can store key-value metadata—essential for grouping, linking, or adding attributes for downstream optimization.

  • If a metadata key starts with _ (underscore), it can be used to group several outcomes together.
    Example: For a “layout” experiment, use _title, _subtitle, _font, _color as metadata to link everything under a single test group.

• CSS Experiments:
  You may add multiple CSS rules to a single outcome for full flexibility. Only applies to experiments of type CSS.

Outcome Naming

For experiments, "name" is for internal reference; for outcomes, the "value" is what matters.

• If the value is self-explanatory (e.g., "red"), there’s no need for a separate name.
• If the value is opaque (e.g., a color code like "C6AFB1"), use the name to clarify ("Rose Quartz").

---

Need advice on complex experiments, grouping outcomes, or CSS test logic?
Contact support here. — our team can help with advanced configuration and optimization strategy.