Wargaming Powers Up Player Happiness with Unified Data and Faster Insights

Twenty-five years ago in Minsk, Belarus, a small group of nuclear physicists with a passion for video games started making their own titles. Their breakthrough came in 2010 with World of Tanks, which they released on a novel “free-to-play” structure: Players could access and play the game for free, with optional in-game purchases that enhanced their experience.

A radical departure from the traditional pay-to-play structure of big-box titles, Wargaming’s player-first principle helped World of Tanks soar to success, reaching over 200 million players around the world. Today, the free-to-play model is widely adopted across the industry, and Wargaming continues to redefine the global video game landscape with over 15 titles.

The award-winning global gaming powerhouse analyzes vast amounts of data to level up game design, personalize player experiences, and refine monetization strategies. To streamline data management, Wargaming migrated from Hadoop and Google BigQuery to Snowflake, unifying multiple data sources into a single platform. This drove significant cost and time savings, accelerating data-driven decision-making to deliver on their vision “to make every player happy.”

Beyond standard game stats, Wargaming tracks player positioning, shot accuracy and movement patterns to deliver the most engaging experiences. If a player struggles at a specific tier, for example, they may receive tailored recommendations, such as missions to complete or recommended purchases. In addition to sentiment analysis, competitive intelligence and user research, Wargaming uses advanced analytics to predict and prevent churn.

The company originally built an on-premises data warehouse with Hadoop, Oracle, Cloudera, Tableau and Google BigQuery. But a growing player base and expanding data volumes exposed this setup’s limitations. Rising storage and processing costs, along with expensive hardware upgrades, made on-premises systems and Google BigQuery unsustainable. Hadoop Impala’s resource constraints required complex ETL workarounds, while hardware expansion — for example, adding six new nodes — could take up to six months, slowing Wargaming’s ability to scale.

The company required a robust platform that could not only consolidate its diverse data into a single, unified location but support emerging initiatives, such as its expanding machine learning models and data science projects. To meet these demands, Wargaming turned to Snowflake for a scalable solution that could provide the flexibility and performance required to power the company’s next phase of growth.

Wargaming explored a range of alternatives to keep up with the fast pace of gaming, including Google Cloud, Amazon and Microsoft Azure — but Snowflake was the clear winner. 

“We found that Google BigQuery and Amazon Athena couldn’t match Snowflake’s superior performance as an analytical query processing engine and platform,” says Alex Ryabov, Wargaming’s Head of Data Services. The team was also attracted to Snowflake’s cost efficiency and real-time visibility into spending, with up-to-the-minute price estimations.

Another advantage was Snowflake’s platform-agnostic storage, which allowed Wargaming to store data anywhere. “Unlike Google, Amazon or Microsoft, which tie you to their ecosystem, Snowflake gives us the freedom and flexibility to choose any cloud provider, or store data in open table formats like Iceberg Tables,” says Ryabov.

Wargaming migrated a total of 644 tables, one petabyte of data, 130 dashboards and over 2,000 data ingestion pipelines; they moved raw data first, then reengineered aggregations and dashboards in Snowflake. This allowed them to avoid the complexity of lifting and shifting old ETL pipelines, which had become a bottleneck in previous architecture. “Throughout the migration, we’ve had exceptional support from the Snowflake team,” says Ryabov. “They proactively check in to keep us on track, and it’s been a fantastic experience.”

Snowflake’s ease of use helps teams get up and running quickly. In particular, Snowpark gives data engineers the flexibility to use their language of choice — in Wargaming’s case, Python — to more easily build data marts for their data scientists. 

Wargaming has also benefited from robust built-in governance, including role-based access control through the Snowflake Horizon Catalog. “Governance is critical, especially with a team spread across different regions and departments,” Ryabov says. “Snowflake’s role-based access ensures the right people get access to the right data at the right time, streamlining workflows and reducing security risks.”

By migrating to Snowflake, Wargaming has reaped substantial time and cost savings. “By reducing our Google Cloud expenses without increasing our Snowflake costs, we save a significant amount, all while staying within budget,” Ryabov says. “Snowflake’s ability to unify data and consolidate workloads into a single platform has greatly boosted our efficiency.” In total, Wargaming saves approximately $500,000 annually, eliminating the need for expensive server replacements and monthly maintenance, and optimizing compute resources.

The move to Snowflake has resulted in an up to 40% increase in data delivery efficiency, giving product analytics teams faster access to near real-time insights through Tableau dashboards. This speed is essential for responding to the fast-moving demands of game development and feature releases. “It’s not just about the speed — it’s a smoother experience overall with Snowflake,” says Ryabov. 

The efficiencies gained with Snowflake have also allowed Wargaming to handle its growing data — 30 terabytes per month — without needing to expand its team. This has freed up time to focus on more strategic initiatives, including faster feature development and optimizations that directly enhance the player experience.

Before the migration, the company’s old system, which used Impala as the querying engine, was notoriously resource-intensive. This caused delays during ETL processes, with analysts often facing frozen screens and query timeouts. The shift to Snowflake has alleviated these frustrations: The centralized environment eliminates silos, boosts performance, and enables faster, more comprehensive insights through unified dashboards and analytics. 

“We used to have data scattered across Hadoop, Google Cloud and other systems,” says Ryabov. “With everything in Snowflake, our team saves time and can deliver more insightful, comprehensive reports.”

As Wargaming looks toward the future, the company is poised to explore new gaming technology, with a strong focus on AI and data optimization. “Generative AI is the next revolution in gaming and business,” Ryabov says. “We’re building an LLM platform and developing chatbots to engage players in new ways, starting with a proof of concept using Snowflake Cortex AI.”

Plus, with Snowflake available across various regions, Wargaming is able to unify its data on a global scale while ensuring compliance with local privacy laws — a significant benefit for a company with an expansive, international player base.

“Next year, our primary focus will be on optimizations: getting everyone comfortable with the platform, stabilizing our data infrastructure, and refining our data ingestion, aggregation and visualization processes,” says Ryabov. “Snowflake will provide immense opportunities for Wargaming’s growth and help us deliver on the cornerstone of our success: player happiness.”