Analytics Integration

Connect Google Analytics to unlock two powerful features:

1. Use GA4 events as conversion goals - Track any GA4 event (purchases, signups, etc.) as your experiment goal
2. Measure Dalton's overall impact - Compare Optimized vs. Control traffic in your GA4 reports

Learn more about Conversion Goals

Events Sent to Google Analytics

Dalton sends two events to GA4 on every page visit where an experiment is active:

Event	Parameter	Values	Purpose
dalton	type	optimized or control	Identifies whether the visitor is in the Optimized or Control group
dalton_experiment	variantId	The active variant's ID	Identifies which specific experiment variant the visitor is exposed to

Use the dalton event to measure Dalton's overall impact (Optimized vs. Control), and the dalton_experiment event to break results down by individual experiment.

Always pushed to the data layer

These events are pushed to window.dataLayer on every qualifying page visit, regardless of whether you've connected GA4 to Dalton. That means you can pick them up in GTM (or any other data-layer-aware tool) even before completing the GA4 integration.

Use dalton_experiment carefully for raw analysis

The dalton_experiment event is useful if you want to pull raw exposure data into your own analysis. Be aware though that interpreting it directly is risky: Dalton uses moving traffic splits and multi-armed bandit allocation, so the share of traffic going to each variant changes over time. A naive conversion-rate comparison in GA4 will conflate variant performance with allocation shifts and seasonality.

Dalton's dashboards correct for adaptive allocation and seasonality to produce a robust uplift estimate — for decision-making, rely on those numbers rather than raw GA4 splits.

Using Google Tag Manager

If your GA4 setup is wired through Google Tag Manager, the type and variantId parameters won't reach GA4 unless you register them in GTM.

1. In GTM, create two Data Layer Variables:
   • Variable name: type → Data Layer Variable Name: type
   • Variable name: variantId → Data Layer Variable Name: variantId
2. Open your GA4 Event tag (the one forwarding events to GA4).
3. Under Event Parameters, add:
   • type →
   • variantId →
4. In GA4, register type and variantId as Custom Dimensions (Admin → Custom definitions) so they become available in reports and audience builders.

Custom events blocked in GTM?

Some GTM setups restrict which custom events are allowed to forward to GA4. If the dalton and dalton_experiment events aren't showing up in GA4, you may need to explicitly add them to your allowed-events list (or create a dedicated GA4 Event tag triggered on event equals dalton / dalton_experiment).

Holdout Group for Impact Tracking

Dalton uses a holdout group to measure overall performance:

• Optimized (80%): Traffic seeing Dalton's experiments
• Control (20%): Traffic seeing original pages (no experiments)

Compare these two groups in GA4 to see the combined effect of all Dalton optimizations.

Adjustable Holdout

The control percentage (default 20%) can be changed in Dalton settings.

Setup in Google Analytics

Follow these steps to set up audiences in GA4 that track Optimized vs. Control traffic.

Step 1: Create New Audience

Navigate to Settings → Audiences → New Audience

Step 2: Configure Custom Audience

1. Click Create a custom audience
2. Select Events as source
3. Choose dalton event

Step 3: Set Audience Conditions

Create two audiences using the type dimension:

• One where type = 'optimized'
• One where type = 'control'

View Results in GA4

Once audiences are set up, compare Optimized vs. Control performance in any GA4 report.

Add Comparison

1. Open any report (e.g., Landing Pages)
2. Click Add Comparison at the top

3. Create comparison groups based on Audience Name

You'll now see side-by-side performance of Optimized (Dalton experiments) vs. Control (original pages) for metrics like conversion rate, bounce rate, and revenue.