How Kindred Optimized its Product Analytics Infrastructure & Costs with Snowplow

Snowplow has been a catalyst for positive change at Kindred. By providing our data analysts with access to granular behavioral data and self-serve data modeling tools, we’ve advanced beyond last-click attribution and can now understand detailed customer journeys. Further, our Product Analytics team is the first to be fully operational in our cloud data warehouse, paving the path for other teams to follow. We’re looking forward to leveraging Snowplow to unlock additional use cases such as dynamic content generation and personalized product recommendations.”

LESYA LISKEVYCH | HEAD OF PRODUCT ANALYTICS, KINDRED GROUP

Background

Kindred Group is one of the world’s leading online gambling providers, home to nine brands with 3 million active users across Europe and Australia in 2023. Among its well-known brands are Unibet and 32Red. Holding six top five market positions, Kindred strives to be a leader in all segments and sets high standards for responsible gaming with its ‘Journey to Zero’ initiative. Kindred is committed to offering thrilling and safe entertainment experiences across gaming, sports betting, poker, and other categories.

In pursuit of its strategic vision, Kindred Group aims to enhance its analytics capabilities and delve deeper into customer journey analysis. However, it recognized the need to evolve its technology to achieve this. While the previous system met earlier needs, it faced limitations as the business grew and evolved. Kindred encountered challenges in consolidating behavioral data, conducting comprehensive data modeling, and managing the associated cost increases. As data collection expanded for more insightful customer and product analysis, these constraints became increasingly apparent.

After swiftly implementing Snowplow’s Customer Data Infrastructure (CDI) in just eight weeks, Kindred Group successfully overcame these hurdles, unlocking significant value. By consolidating, validating, enriching, and modeling all customer behavioral data with Snowplow, Kindred now processes over 2 billion quarterly events with an impressive 97% accuracy in its cloud data warehouse. This new capability allows the Analytics teams to focus on understanding and improving the customer experience and cements Kindred Group’s position as a leader in the betting and gaming industry.

Challenges

Kindred Group initially relied on Tealium for behavioral data collection, Adobe Analytics for digital analytics, and Optimizely for A/B testing. As the business grew, so did the costs and the need to derive more value from the collected data. The cost of scaling these tools further didn’t make economic sense, and accessing the necessary data from Adobe Analytics was becoming increasingly difficult.

Kindred Group subsequently launched an initiative to transition its data from Adobe to its Oracle data warehouse. However, this initiative proved to be both an operational challenge and a financial burden. Lesya Liskevych, Head of Product Analytics at Kindred Group, noted the difficulties:

When we attempted to integrate data from Adobe Analytics into our existing platform, we encountered significant processing and performance issues.”

LESYA LISKEVYCH | HEAD OF PRODUCT ANALYTICS, KINDRED GROUP

Despite the team’s year-long effort to migrate data from Adobe to Oracle, the initiative ultimately faltered due to the volume of events and the inability of the Oracle data warehouse’s to effectively manage this type of data. These obstacles restricted access to the granular behavioral data Kindred Group needed to understand and improve customer journeys.

Solution

Optimizing Analytics Infrastructure: Kindred Group’s Transition to Snowplow

To tackle its challenges and advance its product analytics ambitions, Kindred Group made the strategic shift from Adobe Analytics to Snowplow.

Snowplow provides Kindred’s Product Analytics team with greater modeing flexibility and much more granular data access within the company’s AWS Redshift cloud data warehouse, while optimizing costs.

As Liskevych explains:

Our business case for adopting Snowplow was to optimize costs, as well as improve scalability and efficiency compared to the previous technology.”

LESYA LISKEVYCH | HEAD OF PRODUCT ANALYTICS, KINDRED GROUP

Kindred implemented Snowplow in less than two months by focusing upfront on defining  the intended business outcomes and the behavioral data needed to achieve those outcomes. This preparation enabled their developers to quickly create customized event definitions and schemas, resulting in a rapid deployment of the tracking.

Liskevych states:

We went live with Snowplow in just eight weeks, exceeding all expectations.”

LESYA LISKEVYCH | HEAD OF PRODUCT ANALYTICS, KINDRED GROUP

She credits much of this success to Snowplow’s outstanding customer support: “The service we received was unparalleled.”

Moreover, Kindred’s adoption of Snowplow expedited the company’s migration to the cloud: “Our Product Analytics team is pioneering Redshift and other Amazon services for comprehensive product analyses,” Liskevych explains.

Results

Reduced Costs & Improved Scalability

With the implementation of Snowplow, Kindred has streamlined its operations and significantly reduced costs compared to its previous technology stack. These changes have resulted in an 18% reduction in infrastructure costs, including AWS processing and data transformation costs, while increasing data volumes by 40%.

Snowplow, serving as a core behavioural data engine for analytics and marketing data applications, has simplified the integration effort for the engineering team. This reduction in the need for multiple vendor solutions has freed up time and resources, allowing the team to focus on developing more in-house analytics functionality.

Beyond the cost savings, Snowplow offers Kindred the scalability required to seamlessly manage its vast data volumes. Liskevych underscores the scale, stating:

We process 2.1 billion events through Snowplow in one quarter.”

LESYA LISKEVYCH | HEAD OF PRODUCT ANALYTICS, KINDRED GROUP

This reliable scalability provides a strong foundation for future growth.

Deeper Insights into the Customer Journey

Querying customer data has become much easier for Kindred. Analysts can now access deeper insights into the customer journey faster than ever before. By using Snowplow’s self-serve dbt modeling, Kindred has accelerated the generation of these insights. With dbt and AWS Redshift, analysts can build models on their own and no longer need to rely on the central data team. This autonomy enables the Product Analytics team to swiftly gain a comprehensive view of the customer journey.

Additionally, Kindred Group saves time on maintaining data accuracy. Snowplow’s automated data validation feature notifies the team of errors in real time, boosting data accuracy to 97% with minimal effort.

Unlocked Use Cases for Analytics & Marketing

With granular, structured behavioral data in the cloud data warehouse, Kindred has gained the flexibility to rewrite attribution models, moving beyond the last-touch model to understand the entire customer journey. Liskevych emphasizes this benefit:

Having access to structured atomic data from Snowplow means we can build attribution however we like.”

LESYA LISKEVYCH | HEAD OF PRODUCT ANALYTICS, KINDRED GROUP

This granular data access also enabled the Product Analytics team to launch the “Bringing Customer Journeys to Life” project, which focuses on mapping complete customer journeys. The combination of multi-touch attribution with comprehensive customer journey mapping has helped Kindred to inject customer-centricity into its product experience.

Enhanced Data Governance

Snowplow has enabled Kindred to consolidate its multitude of vendors into a unified cloud data platform on AWS, enabling centralized governance for all event flows, and eliminating siloed third-party tools.

At Kindred, schemas play a crucial role in data governance by defining the structure of collected data to ensure compliance with predefined formats and validation criteria. This systematic approach streamlines data processing by validating only the events that match the schemas through the Snowplow pipeline. This ensures data cleanliness and consistency, while segregating invalid events, strengthening the integrity and reliability of Kindred’s data governance.

Looking Forward

Kindred’s robust behavioral data foundation enables the company to stand out in a competitive market. By expanding its experimentation program with open source tools and Snowplow, Kindred aims to deliver exceptional customer experiences. Initiatives like multi-armed bandit A/B testing for optimizing gaming and sports content and building semantic knowledge systems demonstrate Kindred’s commitment to innovation and customer satisfaction, shaping the future of online gaming.

‍