How Cloud‑Native Data Sharing with Snowflake Is Redefining Data Warehousing

Cloud‑native data warehouses such as Snowflake are changing the way data warehouses and BI systems are built and thought about.

Snowflake’s core feature is data sharing, which allows every Snowflake user to securely access any other database, even across different companies, with personalized views and without copying, preparing, extracting, downloading, or transferring data files.

Developers can treat Snowflake as a single global database; each user operates on the same database but with access rules that protect data confidentiality.

By using simple SQL commands or GUI clicks, these rules can be changed to enable Snowflake’s data sharing.

When one user shares data with another, they grant SELECT or other permissions on existing tables. In a traditional database this is possible within the same instance, but with a shared global instance anyone can share data with anyone else.

Many data warehouses contain internal company data, partner data, supplier data, and paid data. For example, consumer‑goods manufacturers rely on sales data from retailers; without retailer data they cannot understand who bought what, where, and when.

Before data sharing, manufacturers had to extract data from retailers, transfer it to the CPG company, and then load it into another database, incurring significant cost, delay, inefficiency, and loss of low‑level analytical capabilities.

The complexity of big‑data processing often forces even the largest retailers, such as Walmart, to rely on technology vendors to manage their data warehouses and BI systems.

With data sharing, there is no need to extract, transfer, load, or maintain anything; the data owner (the sharing provider) simply shares the data, and the consumer runs the relevant queries.

This sharing is free for the provider, while the consumer only needs to execute queries.

Eliminating extraction and transfer costs reduces expenses for software vendors and allows consumers to access fresher, more detailed records.

However, responsibilities shift to the BI tool layer. Without an ETL process, there is no place to transform inbound data, requiring seamless connection between an organization’s data and another’s. Using shared data as a source for further transformation can re‑introduce copying, which defeats the freshness advantage.

Ideally, BI tools dynamically synchronize data, querying the actual source without storing any copies. In Zoomdata, these capabilities are called multi‑source analytics, originally developed to sync data from Snowflake with Hadoop.

Zoomdata has built extensive features for this purpose, such as cross‑source filtering, data fusion, ad‑hoc grouping, and dynamic synchronization on each query instead of creating static copies.

Future BI tools could help users discover useful shared data, present lists of relevant shared datasets, and provide automated recommendations for machine‑learning, marketing, or trading algorithms.

BI remains the convergence point for business and data. By using databases like Snowflake for data sharing, BI users can leverage both their own data and data from other providers and companies.

Zoomdata envisions BI tools as portals to all world data, allowing users to index, search, and align external data with their own with just a few clicks.