Snowflake vs. Databricks: Who will win the data war?

Everyone loves a rivalry. From the feud between Shakespeare’s Capulets and Montagues, to the ‘El Clásico’ competition between Real Madrid and Barcelona.

Rivalries do more than just create drama and attention. They can also be a very powerful catalyst for motivating teams to work harder, and a way to get people emotionally invested in one side over another. That does wonders for cementing loyalty in a client and partner base.

The technology industry has had its fair share of rivalries. If you’ve been in the tech sector for as long as I have, you might remember the PC war between Apple and Microsoft, or the browser war between Netscape and Microsoft, or the mobile war between Apple and Google, or the CRM war between Oracle and Salesforce.

Tech’s newest rivalry is between Snowflake and Databricks – a rivalry that was on full display last month in San Francisco, where both held their user conferences within days of each other drawing crowds of 16,000+ at each.

It brought back vivid memories of Salesforce and Oracle duking it out in the cloud’s first wave with Dreamforce and Oracle Open World, albeit smaller with shorter keynotes and less sailboats and Buddhist monks. One could almost imagine Snowflake the Bear and SaaSy hanging out at the Moscone show floor.

Snowflake bear

Both Snowflake and Databricks invited Nvidia CEO, Jensen Huang, to join their keynote for a mutual love fest, and both not surprisingly placed AI front and center. While Snowflake leaned into a theme of “Enterprise AI”, Databricks leaned into “Data Intelligence for All”. Both highlighted the growth and importance of their partner ecosystems.

While Databricks and Snowflake often coexist inside customers, the growing rivalry between the two is forcing data and AI consulting partners to make a choice between which platform they should bet their business on.

Both Snowflake and Databricks are courting service providers to serve their growing customer base, and while the two data platforms often coexist inside large customers, the growing rivalry between the companies is forcing data and AI consulting partners to make a choice and decide which of these ISVs to bet their business on.

Evolution of the Snowflake and Databricks rivalry

Tercera has been tracking the Snowflake and Databricks partner ecosystems as part of our Tercera 30 research since its inception, and the growth that these two firms have achieved in the last 10 years has been more than impressive.

Snowflake gained market traction as a cloud data warehouse that addressed the limitations of traditional on-premise data warehouses like Oracle and Teradata that are optimized for structured data (data organized into a standard format with clearly defined data attributes). Snowflake adoption took off due to its ease of use and performance. In September 2020, Snowflake had what is still the largest IPO ever by a software company, raising $3.4 billion at the time. The company now reports more than $2.6 billion in annual recurring revenue (ARR) and more than 9,000 customers.

Databricks was founded by the creators of Apache Spark, an open source distributed computing system known for processing large datasets at speed. Databricks was designed to integrate big data tools in a way that could help teams extract business intelligence from data in a cost efficient way. Optimized more for unstructured data like text, images and videos that don’t have a predefined format, Databricks is seen as having a leg up on Snowflake in the GenAI gold rush. The company is indeed growing faster than Snowflake, albeit on a smaller revenue base (which is estimated to be $1.6 billion in FY24 ARR).

Below is a June 2024 summary from Cleveland Research comparing growth between the two companies.

Today, business and functional buyers within companies tend to prefer Snowflake, while data scientists and data engineers lean into Databricks. However, that is starting to change as the two companies rapidly evolve their offerings into each other’s sweet spots.

Snowflake continues to advance its capabilities in unstructured data with offerings like Snowflake Cortex AI and Snowpark. On the flip side, one of Databricks’ fastest growing products is Databricks SQL, its serverless data warehouse for storing structured data, which Cleveland Research cites to be 25% of revenue and used by 4,500 customers.

Both companies are also pushing hard into interoperability between different table formats and compute workloads. This should make customers’ lives easier and help alleviate concerns around vendor lock-in, which should drive help unlock sales cycles and open up the addressable market for each of the vendors. However, the ability to more easily move data in and out of systems means these vendors need to do everything in their power to keep customers happy, which should fuel the rivalry even further.

A tale of two partner ecosystems

Both Snowflake and Databricks understand that service partners are necessary to help their growing customer base successfully migrate, optimize, use and monetize the data on their platforms. Especially in a consumption-based revenue model, which means the vendors only get paid for what customers use on and around their platforms.

Comparison of the Databricks and Snowflake ecosystem

Source: Houlihan Lokey, Data Science and Analytics Industry Overview, June 2024

Both vendors have invested heavily in their partner ecosystems over the last few years, hiring alliance executives from large SaaS players like Salesforce who understand what it takes to build a thriving partner community. Both have built out online marketplaces, and both are wooing the Global Systems Integrators and trying to expand the number boutique consultancies specializing in their platforms.

However, Snowflake’s partner program is considered by many to be more structured and mature than Databricks. Snowflake has far fewer partners than Databricks, with a highly structured, tiered program that makes it easier for partners to stand out. The company emphasizes deep vertical expertise and encourages its field team to co-sell with partners into large enterprise accounts.

In contrast, Databricks remains more horizontally focused at the moment and has a larger, more fragmented ecosystem. The partner co-selling motion is still a work in progress in the field. However, Databricks is rapidly evolving their program, even making important capital investments to support and build services partners – something Snowflake seems resistant to do (so far).

So who will win?

Tercera believes that both the Snowflake and Databricks ecosystems offer huge potential for up-and-coming IT services firms, and that the global data and analytics market is big enough for both companies to thrive. The interest and adoption of GenAI is creating massive tailwinds for data-related technologies and services, and different customers will have different needs.

It’s not just Databricks and Snowflake who understand what a lucrative market this is. Cloud vendors are investing heavily here too, with platforms like Microsoft Fabric, Salesforce Data Cloud and SAP Datasphere starting to gain traction. Some are more complementary and play nicer in the same sandbox than others.

Consulting partners will need to determine if and how they weave these different platforms into their go-to-market strategy and service offerings — where they specialize and where they diversify. For example, Zennify is finding success with Salesforce Data Cloud and Databricks among its banking and insurance clients. Hakkoda – a dedicated Snowflake partner – is also partnering with AWS for certain healthcare, financial services and public sector clients. (Disclosure: both are Tercera investments.)

The true winner of this rivalry won’t necessarily be any of these companies but the innovation and advancements that result from it. Just as the rivalry between Apple and Microsoft spurred growth and improvements in personal computing, so too will the competition between these companies drive advancements in data management and AI.

Share this: