Powered by Snowflake: Building a Connected Application for Growth and Scale

A major change is coming to how B2B SaaS applications are built. Industry luminaries and experts are seeing it—as signaled by venture capitalist Patrick Chase, Principal at Redpoint Ventures—and we’re seeing it with our Powered by Snowflake partners and enterprise customers. It’s a shift toward connected applications that put customers in control of their data.

This is the first part of a two-part blog series on connected applications. In this post, we’ll explore the connected-application model and why you should consider it both as an application provider and as a customer. In the next blog post, we’ll review how connected applications work and how they are built.

Until recently, SaaS application providers invested in big data capabilities that vastly outperformed what their customers could run in their own environments. As a result, their applications needed to ingest their customers’ data so it could be processed within the SaaS app providers’ own, highly performant data platform. This required customers to relinquish control of their data and to build and manage API pipelines for each of their SaaS applications in order to access the data. We call apps that implement this traditional model managed applications because the app provider manages the data platform where customer data is stored. 

With the broad availability of the Snowflake Data Cloud^™ and other modern data platforms, enterprises of all sizes are adopting massively scalable and cost-efficient data platforms that are on par with those of their vendors. SaaS providers can now store and process their customers’ data on the customers’ data platforms rather than on their own. We call these applications connected applications because SaaS providers connect to a customers’ data platform instead of loading data into their own managed data platform. 

Decision-makers in finance, marketing, and security teams are choosing to implement SaaS solutions that don’t create data silos and that keep their company in control of its data. The key to realizing these benefits is selecting SaaS applications that separate SaaS code from SaaS data. 

When developers build applications on Snowflake^® or other modern data platforms, they can separate their code from the customer’s data in such a way that the customers retain control of their data, while developers create and maintain the code—creating connected applications.

While Snowflake’s unique architecture has always offered significant technological advantages to application developers, an increasing number of Powered by Snowflake partners are realizing the connected-app model can help them further differentiate their products, increase value for customers, and improve go-to-market alignment with Snowflake.

Here are six reasons why our customers are choosing connected apps that tie directly to their Snowflake instance:

1. Connected apps enable a single source of truth across the enterprise

The explosion of SaaS applications has led to a high degree of fragmentation. According to a recent study by Okta, their customers with over 2,000 employees maintain an average of 175 applications! Because most SaaS applications manage their customers’ data in the traditional managed model, each application effectively creates a new silo, as illustrated in Figure 1. This leads to a fragmented data landscape for customers and holds back their ability to become data-driven. 

Figure 1: In a traditional SaaS application approach, each application holds customer data separately.

In the connected-app model, the data is instead centralized by the customer and each app connects to the customer’s data platform (see Figure 2). This prevents silos and puts customers in control of their data.

Figure 2: In a connected-app approach, app providers connect to the customer’s data platform.

2. Connected apps prevent an API integration backlog

Many application providers that don’t implement the connected-app model provide access to customer data via an API. This puts the burden on the customer to build and maintain API pipelines for each application. Most companies have many SaaS applications, which means multiple pipelines, long integration backlogs, and data sets that are never centralized. There’s no need to build and maintain API integrations with connected apps, because the customers’ application data lives in their Snowflake instance. Note also that these concerns do not apply if the managed application is built on Snowflake because in that case, customers can easily access their data using Snowflake’s Secure Data Sharing feature rather than via an API.

3. Connected apps enable custom analytics

With data from all applications centralized in their Snowflake instance, business users can easily analyze data sets across applications. For example, a compliance analyst might want to correlate termination records in the HR application and authentication events in the identity management service. A marketing manager could target an offer by analyzing support ticket data from the customer service application along with customer purchase history from the CRM. These types of cross-application analytics would be difficult to perform across API interfaces. 

4. Customers can apply their own data governance policies

By keeping their apps’ data in their own Snowflake instance, customers can ensure that their data governance policies are consistently applied. For example, some vendors delete records, such as access logs and financial transactions, from their systems after a few months. With the connected-app model, customers no longer have to work around their application providers’ limited retention windows and can flexibly determine their own.

5. Connected apps give slow SaaS a speed boost

Customers that struggle with the performance of a SaaS application can do something about that with a connected-app model. When customers are in control of the Snowflake instance, they can easily scale up the Snowflake warehouse that the application provider is using, thereby speeding up searches and dashboards for their application. This boost can be temporary, for example, just around the busy audit season, or permanent as the company grows.

6. Connected apps offer actionable reporting across the enterprise

Most SaaS applications offer some reporting ability, but their dashboards tend to be basic— they don’t provide access to all the data, and they require you to sign into the application. If the data lives in the customer’s data platform, it can be used in enterprise BI reports with standard tools such as Tableau that can be customized to their needs and accessed by everyone in the company. There’s no more need to export CSV files or send screenshots.

Customers aren’t the only ones who benefit from the connected-app model—so do SaaS application providers. Here’s what we’ve learned from our experience with our Powered by Snowflake Partners.

Application providers like the connected-app model because their customers love it, enabling them to differentiate their application. Customers love having control of their own data and often prefer the connected-app model from vendors that support it. Being able to respond to this preference gives app providers an edge and the chance to win more business. Here are several reasons why customers like 

The connected-app model enables stronger collaboration with Snowflake. Snowflake has a consumption-based sales model and Snowflake sales reps are always looking for new use cases and workloads that can provide customer value and drive consumption for Snowflake’s 5,400+ customers (as of October 31, 2021). As a result, application providers that offer the connected-app model can benefit from collaboration with their account team and a greater opportunity to market their applications to Snowflake’s customer base.

Prospective customers often raise objections related to compliance and security when selecting new SaaS applications that store their data in the provider’s data platform.  For example, many customers have specific data residency requirements such that their data must be stored in a particular cloud region. That’s not a problem in a connected-app model, because the data remains in whatever region(s) the customer has chosen. Other prospective customers may have concerns about the maturity of a startup’s security, which can also be addressed with this model. By implementing the connected-app model, application providers can avoid such concerns, reducing friction from their sales cycles.

Building on Snowflake helps app providers protect their margins because of Snowflake’s consumption-based pricing model: Application providers pay for compute by the second, and they pay only for what they use. The connected-app approach improves application providers’ margins further because the Snowflake consumption and cost is on their customers’ systems, not on their own.

Although there are some new challenges for app providers who implement the connected-app model, we believe that the advantages mentioned above outweigh any disadvantages. For example, some partners have shared that while supporting cross-customer analytics for internal optimizations is possible within the connected-app model, it is more easily implemented within the managed app model.

The connected-app and managed-app models are complementary and not mutually exclusive. Application providers that build applications powered by Snowflake should consider supporting both models. In this way, they can deploy the connected-app model for customers who have or can open a Snowflake account and the managed app model for other customers, giving them the flexibility to meet different customer needs and increasing their addressable market. Powered by Snowflake partners such as Panther and Hunters support both models.

A great example of the benefits of the connected-application model is in the cybersecurity space. Historically, security teams collected their data in dedicated Security Information and Event Management (SIEM) solutions. These systems would operate separately from the rest of the enterprise’s data stack, with painful limits on scale and flexibility. As a result, most security organizations contended with numerous data silos isolated from the rest of the business.

To remove these limitations and meet the security needs of NETGEAR’S fast-growing cloud presence, Pallavi Damle, Head of Global Cybersecurity at NETGEAR, turned to a connected app on the Data Cloud. As described in this blog post, Damle achieved her visibility and automation goals, along with significant cost savings, by moving from a dedicated SIEM solution to running security analytics alongside NETGEAR’S other analytics workloads on Snowflake. This shift was enabled by the SaaS solution that NETGEAR purchased from Hunters. Hunters, a Powered by Snowflake partner and Snowflake Ventures portfolio company, provides threat detection and response capabilities that customers can run standalone or on top of their existing Snowflake instance.

Hunters loads security data into NETGEAR’S Snowflake account and powers its interface and automation with NETGEAR’S Snowflake compute resources. The direct access to NETGEAR’S Snowflake resources means that Hunters’ threat detection engine taps into and combines more complete and diverse data sets. If there’s a security incident that requires rapid investigation, the NETGEAR information security team can quickly scale query power as needed because NETGEAR is in control of the warehouse size. It’s an attractive model, and it played a key role in NETGEAR becoming a Hunters customer.

Hunters is not alone. Exciting security industry players such as Securonix and Panther are also adopting this innovative approach.

The increasing popularity of the Snowflake Data Cloud and other modern data platforms is causing many enterprises to choose the connected-app approach for their SaaS applications. Enterprises retain control of their data while SaaS providers enjoy new growth opportunities. Importantly, SaaS providers that support Snowflake benefit from a strong go-to-market alignment with Snowflake sales teams and potential access to thousands of Snowflake customers. 

What’s more, integration with Snowflake is very straightforward. There is usually no need to engage in a lengthy re-platforming project. Snowflake-provided connectors and services can be quickly leveraged for bidirectional integration, resulting in very attractive ROI for SaaS providers.
Stay tuned for our next connected-apps blog post, where we’ll explore how these apps work and how you can build them. If you’re looking for more information about how you can create apps that separate code and data, become a Snowflake partner through the Powered by Snowflake program. The Powered by Snowflake program guides members through designing and implementing applications that can run directly on a customer’s Snowflake instance. Once you launch your solution, the program also helps you raise awareness of your solution among Data Cloud users.