تابعنا على
Delivery Hero Replaces Google Analytics with In-House Tracking Platform

Dev News

Delivery Hero Replaces Google Analytics with In-House Tracking Platform

Delivery Hero Replaces Google Analytics with In-House Tracking Platform

Delivery Hero, a global online food ordering and delivery company, has successfully moved away from Google Analytics and now operates its own internal user tracking platform. The migration was led by Alina Krasavina, who detailed the project in a recent presentation. The new system was built to handle growing data demands while reducing costs associated with third-party analytics tools.

The company developed a scalable architecture designed for simplicity. According to Krasavina, the platform can manage ten times the previous load while capturing 97 percent of tracking data. This performance improvement was achieved without complex infrastructure changes, relying instead on streamlined data pipelines and efficient storage methods.

Google Analytics had been the primary tool for monitoring user behavior across Delivery Hero’s multiple brands and markets. However, as the company expanded, the costs and limitations of the external service became a concern. The internal platform was created to give Delivery Hero full control over data collection, processing, and storage.

The new system processes events in real time, allowing the company to track user interactions with minimal delay. Krasavina noted that the team prioritized data accuracy, ensuring that only a small fraction of events, approximately 3 percent, are lost during transmission. This high capture rate is critical for reliable business analytics and product development decisions.

Architecture and Implementation

The platform uses a queue-based approach to handle incoming data. Events are collected from web and mobile applications, then sent to a message broker before being stored in a data warehouse. This design separates the ingestion process from the storage layer, making it easier to scale each component independently.

Delivery Hero also implemented data validation at the point of collection. Invalid or malformed events are filtered out early, reducing processing overhead. The remaining data is then aggregated and made available for analysis through internal dashboards and reporting tools.

Cost and Performance Benefits

By moving away from Google Analytics, Delivery Hero eliminated licensing fees associated with the third-party service. The internal platform runs on existing cloud infrastructure, which the company already operates for other services. Krasavina said the total cost of ownership for the new system is significantly lower, although exact figures were not disclosed.

The performance gains have allowed the company to expand tracking capabilities without additional investment. The platform currently handles billions of events per month across more than 40 countries. This volume would have been expensive to process using external analytics tools, according to the presentation.

Implications for the Industry

Delivery Hero’s migration reflects a broader trend among technology companies moving away from third-party analytics services. Privacy regulations, data ownership concerns, and rising costs are driving organizations to develop internal alternatives. The approach demonstrated by Delivery Hero may serve as a reference for other companies considering similar transitions.

The project also highlights the importance of scalable data systems in the context of global operations. As businesses collect more user data, the ability to process and analyze it efficiently becomes a competitive advantage. Delivery Hero’s internal platform allows the company to maintain control over its analytics while adapting to changing requirements.

Krasavina emphasized that the migration was completed without disrupting existing workflows. The team ran parallel systems during the transition, ensuring continuity of data collection. The new platform has been fully operational since late 2023, and Delivery Hero continues to refine its features.

Looking ahead, Delivery Hero plans to extend the platform’s capabilities to include advanced machine learning models for user behavior prediction. The company also intends to open source certain components of the architecture, allowing other organizations to benefit from its experience. No timeline for this release has been announced.

Click to Comment

Leave a Reply

Your email address will not be published. Required fields are marked *

More Articles in Dev News