Outcome Overview

Outcomes are the possible variants or results an experiment can serve. How you configure outcome values and metadata directly affects test clarity, grouping, and optimization power.

Key Concepts

• Value and Name:

  • In outcomes, the value is always the primary field.
    • For simple values like red, use the value directly—no name needed.
    • For codes or complex values (e.g., C6AFB1), use a clear name to aid recognition (e.g., “Rose Quartz”).
  • Unlike other admin pages, outcome value is what matters most.

• Metadata:
  Metadata is the centerpiece for grouping, linking, and enhancing outcomes.

  • Use metadata attributes for products, grouping, or complex optimization logic.
  • If a metadata key starts with an underscore (e.g., _title, _font), you can group multiple outcomes into a single test unit.
    • Example: To test an entire layout (font, color, title, subtitle together), use grouping keys like _font, _color across the relevant outcomes.
  • Proper metadata enables advanced reporting and powerful optimizations.

• CSS Experiments:
  For experiments of type CSS, you can define multiple CSS rules per outcome.
  This allows full flexibility in A/B or multivariate UI tests—change colors, layouts, or visibility without any code deploy.

Grouping Outcomes with Metadata

Metadata-driven grouping is essential when testing multiple attributes together. Use underscore-prefixed keys to link outcomes as a set for cleaner reporting and actionable insights.

---

Need best practices for metadata, grouping, or CSS outcome strategies?
Contact support here. — our team can help you unlock more advanced use cases.