Product descriptions:
Sovrn empowers 6,000+ publishers and creators by providing the data and insights needed to thrive on the open internet. To operate at this scale, Sovrn needed to turn siloed Commerce and Ad Exchange data into actionable insights. Previously, fragmented systems, costly infrastructure and complex governance slowed innovation and increased overhead. With the Databricks Data Intelligence Platform, Sovrn unified its data to deliver AI-driven insights efficiently, enabling smarter audience targeting and monetization strategies that increased publisher yield by 30%.
Confronting data silos that stall innovation
Sovrn empowers 6,000+ publishers and creators to thrive on the open internet by delivering the advertising demand that sustains their businesses. Relying on a rich mix of data from its Commerce and Ad Exchanges, Sovrn powers critical business use cases like purchase intent modeling, audience building, traffic shaping, data marketplace initiative and internal financial and operational reporting. With these insights, advertisers can target the right audiences more effectively, and publishers can maximize yield. While Sovrn had the right data, acting on it at speed and scale was challenging.
The company’s data was locked in silos across multiple products and teams, making it difficult to unify and standardize. This slowed product development, limited collaboration across engineering and data science groups and created roadblocks to innovation. Chris George, President of Sovrn explained, “We collect a lot of really great data, but actioning on it has been difficult. Getting everybody on the same platform was crucial for us.”
Technical hurdles compounded these business challenges. Managing EMR clusters was time-consuming and expensive, while Zeppelin notebooks created additional silos that hindered collaboration and productivity. Attempts to modernize with Snowflake added to the complexity, as configuring and maintaining role-based access controls proved costly and unwieldy. Despite all of the data living in AWS, Sovrn lacked a central layer of access and management to maximize its impact. These constraints rippled across Sovrn’s operations. Data science teams struggled to quickly iterate on new models, internal reporting lacked efficiency and transparency into costs across 150+ applications was limited.
Building a governed foundation for scalable growth
Sovrn turned to the Databricks Data Intelligence Platform to unify Commerce, Ad Exchange and operational data on a single platform. This foundation allows engineering, data science and business teams to collaborate seamlessly, speed up product development and deliver insights to customers faster. At the heart of this transformation is Unity Catalog, providing a single, intuitive governance layer across Sovrn’s 30+ AWS accounts. Role-based access controls that once required extensive manual setup now take minutes, giving teams faster, more secure access to the datasets they need. Delta Lake further streamlines management, offering flexibility for reading and writing data without the overhead of traditional formats.
Sovrn also benefits from Databricks’ collaborative workspaces. All-purpose notebooks make it simple to test new ideas and rapidly build proofs of concept without committing to full-scale infrastructure. This agility is key for launching new offerings like purchase intent audiences, which were validated quickly and brought to market under budget. Patrick Browne, Software Engineering Manager for Data Engineering at Sovrn said, “Previously, we’d have to run a compiled Spark job with EMR configurations. Now we can spin up a serverless notebook, aggregate the data, and know right away if there’s something worth pursuing.”
With real-time dashboards, AI and BI tools, and growing exploration of data clean rooms, Sovrn now has a platform that reduces costs, without impacting necessary innovations that opens doors to new ways of delivering value to publishers and advertisers. Through this secured data sharing environment, Sovrn’s vision to break down the “walled gardens” of dominant advertising networks—Google and Facebook—comes to life. “Databricks helps us implement Data Monetization mechanisms through our Marketplace that are imperative when competing for audience attention in the open internet,” explained George.
Reducing spend while accelerating innovation
Since moving to the Databricks Data Intelligence Platform, Sovrn is driving efficiency while creating new opportunities for revenue growth. By migrating from EMR and other legacy systems, Sovrn cut infrastructure costs significantly—achieving a 70% savings that includes $1.2 million in reduced EMR spend. Data engineering teams also gained a 52% productivity boost, freeing time and resources for higher-value work that directly impacts their customers.
Equally important has been the speed of actionable insights. Sovrn’s data pipelines now run 5x faster, allowing teams to process and surface results more quickly. For publishers, this means faster access to audience targeting and performance data that directly impacts yield and campaign outcomes. These efficiencies translate into stronger results for customers. Smarter traffic shaping and audience modeling have delivered a 30% increase in publisher yield, while filtering out low-quality advertising supply saves Sovrn $45,000 per month in EC2 costs. Both publishers and advertisers benefit from higher-quality impressions and more precise audience targeting.
Looking ahead, Sovrn sees Databricks as key to future innovations. AI-driven campaign optimization, secure data collaboration and data clean rooms are opening the door to capabilities that were once impossible. Sovrn is also scaling its data wholesale operations. By aggregating and governing audience data, they are providing actionable insights—such as demographics, purchase intent and browsing behavior—to external platforms like The Trade Desk. This is not only bringing more demand to publishers, but also enabling buyers to better understand their audiences and deliver more effective advertising.
“With a governed data foundation underpinned by Databricks, we can rival the likes of Google and Facebook. Through this marketplace, we can help publishers drive more revenue while supporting a thriving open internet where independent voices can compete,” concluded George.