Skip to main content
All CollectionsAccount & Datasource Set-UpAttribution
How to track custom conversions with Klar
How to track custom conversions with Klar

Now you can send custom conversions to our Pixel and see them inside your attribution dashboards

Marc Garbella avatar
Written by Marc Garbella
Updated over 3 weeks ago

Custom Conversion Tracking in KLAR

Overview

Track on-site conversions and send them to KLAR for in-depth analysis within your attribution reports. Additionally, you can pass values linked to these custom events to measure ROAS and other key performance metrics.


Sending Custom Conversion Events

Event Name

customConversionEvent

Available Fields

  • customConversionId (Integer between 1 and 10)

    • Used to differentiate event types (e.g., 1 = Add to Cart, 2 = Newsletter Sign-Up).

  • customConversionValue (Float)

    • Represents the monetary value of the event (e.g., 20.0).


Example Implementations

This needs to be integrated in your store source code or via GTM.

Shopify Customer Events (Pixel)

Shopify.analytics.publish("Klar:customConversionEvent", { customConversionId: 1, customConversionValue: 20.0 });

Shopify Webpixel Sandbox

api.analytics.publish("Klar:customConversionEvent", { customConversionId: 1, customConversionValue: 20.0 });

Other Pixel Implementations

window._k_q = window._k_q || []; window._k_q.push(["Klar:customConversionEvent", { customConversionId: 1, customConversionValue: 20.0 }]);

Setting Up Custom Conversion Events in KLAR

  1. Navigate to Store Configurator → Open Store Settings.

  2. Define Your Custom Conversions → Assign an ID to a descriptive name (e.g., 1 = Add to Cart), so they appear correctly in the frontend.


Analyzing Your Data

In the Attribution Deep Dive, you'll find detailed insights, including:
✅ Total number of conversions
✅ Conversion values
✅ Cost per action (CPA)
✅ Return on ad spend (ROAS)

Each conversion is attributed to the session in which it occurred. Using KLAR identifiers, we can match conversions down to an ad level for precise attribution.

Did this answer your question?