Skip to main content

Experimentation

EDS sites can add experiment variants with the AEM experimentation plugin. Authors create challenger pages in the same authoring flow, then enable the experiment with metadata on the control page. Measurement uses Real User Monitoring (RUM).

How a test is set up​

Create challenger pages under an /experiments/<experiment-id>/ folder, publish the challengers, then add metadata to the control page:

Metadata rowValue
Experimenthero-test
Experiment Variantshttps://main--site--owner.aem.live/experiments/hero-test/home
  • Experiment - a unique experiment ID used for tracking and reporting.
  • Experiment Variants - one or more published challenger URLs, separated by line breaks when there is more than one challenger.

Variant pages are authored exactly like the original page - same blocks, same authoring flow - just with different content.

Audiences​

Audience and targeting features depend on the experimentation/decisioning plugin and the project's implementation. Do not model targeting as an audiences spreadsheet unless the site has explicitly implemented that contract.

Multivariate​

Add multiple challenger URLs to the Experiment Variants metadata row. By default, traffic is distributed evenly across the control and all challengers. For example, two challengers produce a three-way split: control, challenger A, and challenger B. The plugin supports metadata-based split overrides; keep those overrides close to the control page so authors can review them with the experiment.

Tracking results​

EDS does not write results back into a spreadsheet. The experimentation plugin reports variant exposure through RUM, and you can also wire your analytics tool of choice (Adobe Analytics, GA4, Plausible) from the site's implementation.

Keep custom analytics bridges implementation-specific. Read the selected experiment and variant from the experimentation plugin or data layer used by the site; do not assume a global data-experiment attribute exists unless your project adds one. For example:

scripts/delayed.js
const assignment = window.acmeExperimentAssignment;
if (assignment && window.analytics) {
    window.analytics.track('experiment-exposure', assignment);
}

Stickiness​

The experimentation plugin keeps variant assignment stable for the visitor while the experiment is active. Review the plugin configuration before relying on a specific cookie name or lifetime.

Concluding an experiment​

Two options:

  1. Bake the winner in - replace the original page's content with the winning variant, then remove the experiment metadata.
  2. Promote the variant - delete the original page, redirect to the variant URL.

Option 1 keeps URLs stable and is the default.

Limitations​

  • Variant pages should be published EDS pages, commonly under /experiments/<id>/.
  • Exclude challenger pages from the sitemap and add robots: noindex,nofollow to avoid duplicate-content indexing.
  • Splits are per-experiment, not stacked. If you run two experiments on the same page, visitors are assigned independently to each.
  • Server-side personalisation tied to authenticated identity is out of scope - use the hybrid pattern for those pages.

See also​