How a Leading Automotive Marketplace Drives Real-Time Personalization at Scale

A major U.S.-based automotive marketplace, with millions of monthly users and over 7,000 dealer websites, partnered with RudderStack to streamline behavioral data collection, unify fragmented pipelines, and deliver real-time personalization across its digital properties.

As the company expanded its offerings, from car search to CRM tools, inventory management, and marketing automation, it required a flexible, real-time data infrastructure to support advanced personalization, advertising, and lead generation use cases at scale.

The Challenge: Fragmented Events, Slow Personalization

At the onset, the data stack was fragmented and reactive. Behavioral events from the core marketplace and dealer sites flowed directly into Kafka, but lacked a centralized control layer. Event distribution was handled manually, with batch-heavy pipelines that delayed personalization efforts and gave marketing teams minimal access to behavioral data.

To meet growing demands, the data engineering team needed to build an internal system that could:

• Capture every interaction across both logged-in users and anon visitors.
• Apply enrichment, attribution, and scoring in real time.
• Activate enriched data to drive personalization and acquisition.
• Ensure transparency, flexibility, and compliance.

The Solution: RudderStack as the Core

Rather than implementing a bundled CDP solution, the company chose a composable strategy, placing RudderStack at the heart of its behavioral data infrastructure.

Key Functions:

• Behavioral Event Collection & Routing RudderStack captures granular user events from the core marketplace and 7,000+ dealer-hosted websites. Events are sent to more than a dozen internal and external destinations. Critically, all data is funneled into Kafka for in-stream enrichment, identity stitching, and ML-based scoring.
• Real-Time Transformations + Replayability The company leverages RudderStack’s transformation layer to enrich events as they are ingested, bringing personalization logic closer to the edge. If an integration breaks or a transformation fails, teams can replay raw events from the warehouse without data loss.
• Activation for Personalization and Marketing Enriched events are written into Apache Pinot for low-latency API access, powering real-time personalization based on inferred intent, search activity, and previous views.

“RudderStack gave us a central collection point and made distribution seamless. Adding new integrations is simple, and replayability has saved us multiple times.” — Director of Data Engineering

The Results: From Batch-Based Silos to Real-Time Personalization

RudderStack now sits at the center of the company’s behavioral data stack, replacing a patchwork of brittle, batch-based pipelines with a real-time infrastructure that supports teams across product, engineering, and marketing. Personalization is now dynamic, driven by live behavioral signals that update what users see on each pageview. User profiles combine anonymous and known behaviors, enriched with over 400 attributes updated continuously. Marketing teams can build and sync audiences without engineering support, and push segments into ad platforms in real time. Compliance is handled at the infrastructure level, with centralized PII management and anonymization to meet GDPR and CCPA standards.

Strategic Takeaway: Why Composability Won

The company avoided the lock-in and abstraction of a bundled CDP by building around RudderStack.

Why RudderStack was the right fit:

• Centralized event routing across web, mobile, and backend systems.
• Transformation logic embedded at ingestion time
• Guaranteed delivery and replay capabilities

Today, the organization has turned its behavioral data layer into a strategic asset, driving faster personalization, leaner marketing execution, and new revenue opportunities across its automotive ecosystem.