Unlocking Data Silos with Snowflake’s Secure Data Sharing
In today’s data-driven world, businesses across industries are increasingly realizing the importance of harnessing the power of data to drive informed decision-making. As a leading player in the online video space, JW Player has not only enabled businesses to deliver exceptional video experiences but also recognized the growing demand among its customers to integrate their data directly into their own data warehouses.
Enterprises using JW Player’s robust video platform have vast amounts of valuable data at their disposal. This data encompasses critical information about user engagement, video performance, and more. By integrating this data into their own data warehouses, JW Player customers can consolidate and analyze it alongside their existing business data, leading to a centralized and governed view of their entire operation.
The integration of JW Player data with customers’ data warehouses holds immense potential for empowering businesses to make better and more informed decisions. With all their data in one place, these organizations gain a holistic understanding of their video strategies, customer behaviors, and audience preferences. This centralized view enables them to extract valuable insights, identify trends, and uncover correlations that might have otherwise remained hidden in siloed data sources.
Data Sharing v1
Several years ago, we launched our Data Sharing v1 product, which was built on AWS S3, enabling us to deliver CSV files to dedicated S3 buckets for our customers to access and load into their systems. However, despite its initial promise, we encountered several issues along the way.
We observed that many customers faced difficulties accessing their S3 buckets, particularly if they lacked familiarity with AWS IAM policies. The complexities of configuring access permissions created unnecessary barriers, hindering our customers’ ability to efficiently utilize the shared data. We recognized the need for a more user-friendly approach that would empower our customers to access and leverage the data effortlessly.
Another hurdle encountered with our v1 product was the burden placed on our customers to build and maintain their own ingestion pipelines. After accessing the data files from their S3 buckets, customers were required to establish complex pipelines to load the data into their respective data stores. And it was difficult for our support team to help customers troubleshoot given we had no visibility into these custom workflows. It became clear that a more streamlined and automated solution was necessary to improve the data sharing experience for our customers.
Over time, we observed a decline in the usage of our Data Sharing v1 product. Customer feedback indicated frustration with the accessibility challenges and the complexity of building ingestion pipelines. It was evident that we needed to find a better way to share data with our customers — one that would simplify the process, enhance user experience, and reinvigorate their engagement with our data offerings.
Data Sharing v2
Currently in beta at the time of this publication, our Data Sharing v2 solution leverages a hosted database supported through the robust capabilities of Snowflake Secure Data Sharing.
As a hosted database solution, Snowflake Secure Data Sharing offers a scalable and reliable infrastructure for securely storing and sharing data. Gone are the days of complex configuration and manual processing pipelines. With Data Sharing v2, integrating the shared data into our customer’s preferred business intelligence tool is as simple as adding database credentials. Whether they use popular tools like Tableau, Power BI, or Looker, accessing and querying the shared data has become a breeze.
One of the standout features of Data Sharing v2 is the elimination of the need for developers to build and maintain data processing pipelines. Traditionally, developers would spend valuable time and resources pulling data into their own data stores, creating complexities and potential bottlenecks. However, JW Player customers can now directly query the hosted database without the hassle of additional infrastructure setup.
v2 Implementation
Once a request to enroll a customer in Data Sharing v2 comes in, our job orchestrator Alpaca (our Airflow implementation) gets to work. There are two types of customers that can be onboarded to the product; those who have their own Snowflake account and those who don’t.
If a customer already uses Snowflake, our airflow job enables their Snowflake account access to a share on our account. They can immediately start querying dedicated views in our Snowflake account using compute resources in their own account.
If a customer does not have their own Snowflake account, our airflow job will create a Reader Account within our own Snowflake account. The job enables share access to the customer’s data and customers can query it using compute resources in JW Player’s Snowflake account. We monitor warehouse usage and can keep track of customer compute costs of the product.
Once customer’s are onboarded, they can integrate Snowflake into their own BI tools gaining the advantage of a unified view of all their data, empowering them to meet their specific business needs. And with the product running all in one place, it’s easier for our support team to help customers troubleshoot.
