How Engel & Völkers Powers Smarter Content with Behavioral Data

Background

Engel & Völkers is one of the world's leading companies in luxury real estate brokerage. It operates across residential and commercial properties, as well as yachting and aviation. With over 1,000 locations spanning 30+ countries and more than 16,700 professionals worldwide, the company serves three distinct customer groups: real estate agents, property buyers, and sellers.

As a global luxury brand, Engel & Völkers relies heavily on digital channels for lead generation and customer engagement. Its website features complex user journeys tailored to each customer segment. It is also supported by over 13,000 blog articles across multiple country sites. 

Over 30 people work on the company’s website optimization and support for the franchise network. For this work to be a success, the team requires detailed performance insights to make data-driven decisions. However, Engel & Völkers’ legacy tracking was holding it back from delivering personalized experiences. And crucially, the setup made it difficult to optimize marketing investments. It was this challenge that led the company to Snowplow. 

‍

The Challenge: From Fragmented Logs to Actionable Insights

Before Snowplow, Engel & Völkers operated a DIY analytics pipeline. The setup was built around Elasticsearch, which created significant barriers to understanding customer behavior. 

The system collected events without proper user or session identifiers. As a result, it was impossible for the data team to reconstruct user journeys or track multi-visit behavior patterns. Ivan Serezhkin, Global Head of Data at Engel & Völkers, explained: 

“Data consistency was a nightmare. We had the same field written in 10 different ways across our system - camel case, all caps, lowercase, with underscores, with hyphens. When we tried to report on something as basic as listing ID, data was missing because our BI scripts only recognized some of those formats."
Ivan Serezhkin, Global Head of Data, Engel & Völkers

This fragmented data collection created three major obstacles to growth:

• Data Collection Gaps and Fragmentation: Event tracking was inconsistent, unstandardized, and incomplete. The team couldn’t analyze user paths, funnels, or understand how customers moved through the company’s complex web ecosystem. 
• Lack of Actionable Insights: Engel & Völkers' legacy setup only provided basic, surface-level metrics like lead counts. There was no insight into lead quality, user behavior patterns, or conversion drivers. Consequently, marketing decisions lacked data-driven support. The business couldn't determine where to focus engineering efforts or marketing investment across its diverse customer segments.
• Operational Limitations: The Elasticsearch-based system depended on niche technical knowledge to maintain. It simply wasn't scalable for a luxury real estate business operating across 30+ countries. In addition, the implementation of additional tracking required significant overhead from technology teams, and the process wasn't streamlined. 

These limitations created a huge problem for Engel & Völkers. Its complex business model serves multiple customer types across different geographies. Each segment requires tailored content and user experiences that the company could not deliver with its DIY pipeline. 

‍

The Solution: Building a Modern Customer Data Infrastructure

The Engel & Völkers team chose Snowplow to replace their DIY pipeline with a comprehensive customer data infrastructure built on Google Cloud Platform (GCP). The decision was driven by six key requirements: 

• Full ownership of tracking data without black-box analytics 
• Access to raw enriched event data 
• Real-time streaming capabilities 
• Custom enrichment and identity stitching 
• Seamless integration with their existing GCP/dbt/Tableau stack
• Schema validation for data consistency.

Engel & Völkers’ new architecture now centers around Snowplow’s event tracking pipeline. 

JavaScript trackers are embedded across its website, CRM, and internal tools. These trackers send events to collectors, which flow through an enrichment pipeline where they’re validated against schemas. Following this, clean events are loaded into BigQuery in near real time. In parallel, failed events are flagged for quality monitoring. 

dbt then handles transformations using both Snowplow’s models and custom business logic.  As Muralidhar Reddy Kuluru, Senior Analytics Engineer at Engel & Völkers’ noted: 

"What really helped and accelerated things for us was the official Snowplow dbt repository, the open source one on GitLab. That actually gave us a good starting point because we had pre-built models for sessions, users, and page views, and from there, we could build a lot of custom logic and business logic on top of it."
Muralidhar Reddy Kuluru, Senior Analytics Engineer, Engel & Völkers

Lastly, Tableau provides visualizations, with insights flowing back to the website for dynamic personalization.

The team built an internal tracking library on top of Snowplow that standardizes implementation across different development teams. This library includes schema registry configurations and tracking configs. As a result, the team has ensured consistent data collection even when multiple developers work independently on different website sections. When updates are needed, they're implemented once in the library rather than across multiple tracking implementations.

The team also implemented custom identity stitching logic connecting anonymous users to known customers at conversion points. This has enabled complete user journey reconstruction across devices and sessions. The system handles their complex multi-country, multi-language environment while supporting their franchise network's need for localized insights.

‍

‍

Results: Measurable Impact on User Engagement

A cornerstone of Engel & Völkers' implementation is the smart content recommendation engine. This cutting-edge solution addresses the challenge of surfacing relevant content from the company's 13,000+ blog articles. The system uses an RFM-style scoring model that evaluates each article across three dimensions: popularity (unique users per day), engagement (average time on page), and effectiveness (conversion rate after viewing). These metrics are weighted at 30%, 20%, and 50% respectively to generate recommendation scores, with calculations performed separately for each customer segment to ensure relevance.

The smart content recommendation engine delivered immediate, measurable improvements to user experience and business metrics. The solution surfaces the most relevant blog articles based on behavioral data rather than chronological order. As a result, Engel & Völkers achieved significant engagement gains. As Julian Brandt, Team Lead for Customer Analytics at Engel & Völkers reported:

"We could see a 4% increase in average engagement time on the next page. So users are more engaged in reading content. Interestingly, the drop off rate actually decreased quite a lot. So users actually didn't bounce off our page after they reached this article overview page. And the conversion rate also increased by 5%. So more users convert after consuming recommended content."
Julian Brandt, Team Lead for Customer Analytics, Engel & Völkers

The 30% reduction in drop-off rates demonstrates that users are finding more relevant content, while the 5% conversion rate increase directly impacts lead generation. These improvements validate the team's user-centric approach to content prioritization over traditional chronological sorting.

Beyond the recommendation engine, Snowplow transformed Engel & Völkers' broader analytics capabilities. Snowplow provides more precise data than GA4, enabling detailed marketing optimization and comprehensive reporting.

Raw event access allows effective diagnosis of tracking issues that would be impossible to debug with abstracted analytics interfaces. The team can now analyze complete user journeys from the first touchpoint through CRM activities post-conversion.

The schema registry has evolved how the team approaches tracking, ensuring consistent data structure across applications and teams. This governance eliminates the data quality issues that plagued the company's previous system. Additionally, the separation of event schemas and context schemas enables efficient reuse across different tracking implementations.

‍

Looking Ahead: Expanding Behavioral Intelligence

Building on its content recommendation success, Engel & Völkers plans to apply similar behavioral scoring to property search results. Using user engagement data, the team aims to personalize user search, surfacing the most relevant listings rather than relying solely on traditional filters like price or date. As Julian Brandt noted: 

"That's the next step on using the recommendation scoring algorithm for further enhancing our experience on the website."
Julian Brandt, Team Lead for Customer Analytics, Engel & Völkers

With its new robust data foundation in place, Engel & Völkers is now positioned to continue innovating in behavioral analytics. And crucially, it’s able to do this while maintaining the data quality and governance standards essential for its global operations. 

‍