Increasing SaaS costs and a tightened cost of capital have become a significant driver in purchasing decisions. This week, industry leader Tristan Handy, the CEO of dbt, called curtains on the ‘modern data stack’ hype cycle in his weekly blog, pointing to cost as a major barrier for customers. As Tristan points out,

"As a result of all of this pressure on data spend, buyers developed a strong preference to buy integrated solutions rather than to buy many tools to construct a stack. Buyer willingness to construct a stack from 8-12 vendors has declined significantly. Companies are much more likely today to expect to buy 2-4 products as the core of their analytics infrastructure."

In this post, we’ll dive into how the 'no-longer-modern' data stack accumulates high costs and how Omnata can alleviate this for integration. Snowflake Native Apps open up a new architecture and a different business model deliver more value for customers.

The ‘Modern Data Stack’ stacks up SaaS costs

Think about what goes into a SaaS app under the hood. They are built from the servers up and demand a very high level of engineering to be used by an enterprise. However, in each app, customers pay for the same things;

• Cloud infrastructure and compute resources

• Regional deployments

• Cyber-security

• Performance optimisation

These features are critical. However, they are undifferentiated and only stand-out when they fall short.

SaaS integration has the highest undifferentiated cost

Next, think about the process of moving data between sources and targets. Data is extracted from one cloud, is staged in another, then landed into the final destination. For the longest time, this has been the only way to move data to and from a data warehouse and an ecosystem of SaaS apps spawned to satisfy this use case.

However, customers have been paying too much for standalone middleware to move data.

Usage-based integration pricing double-charges customers

Another staple of SaaS middleware pricing is ‘usage’ based pricing models. ‘Usage’ is usually calculated in data volume like Monthly Active Rows, rows processed or compute time.

Common complaints include

• Costs increase infinitely as data always grows

• High volume use cases must be scaled back

• Growth in cost doesn't reflect the cost to build a connector

• Customers already pay for compute and storage in their data-warehouse

Usage-based pricing can cause some smaller customers to spend as much on moving data to a data warehouse as on the data warehouse itself. Let's be clear, vendors charging usage-based pricing are not price gouging, this model simply reflects the high underlying costs.

Breaking down the underlying costs drivers

In breaking down costs, we'll look at the architecture of a SaaS product which encompasses infrastructure, technology, employees and maintenance. Plus we'll touch on one of the biggest drivers, the cost of capital.

Cost driver: Cyber-security

Researching pricing models from popular SaaS tools, one thing becomes clear. Cyber-security is sold as an upsell to Enterprise plans that can be 2-5x the standard plan.

There is even a website that shames SaaS vendors who upsell SSO. See sso.tax.

But, we're not here to shame anyone! As you'll see, this is less a dodgy pricing strategy as it is a necessary-evil.

Narrowing to integration SaaS, the most common cyber-security features held back are

• SSO

• User and data governance features

• Hybrid-owned data storage

• PrivateLink and advanced connection methods

However, most if not all, of these features need not exist if the tools were not standalone SaaS.

This pricing reflects the underlying cost of building and maintaining SaaS infrastructure and the need for vendors to ‘loss-lead’ on their self-service plans with the eventual goal of moving customers to Enterprise plans.

On the native app model, running on an already trusted platform, this is no longer relevant.

Cost driver: Scaling cloud infrastructure

Beneath every SaaS tool is cloud infrastructure. Vendors pay for both direct cloud bills and also engineering a product for scale. Large cloud customer like Snowflake have huge volume and can negotiate the lowest unit cost. Smaller cloud customer, like most SaaS integration tools, have to limit how their customers can consume the product.

Usage-based pricing is one lever that integration vendors pull to ensure they collect more than they spend, but they also set soft limits. For example, every cloud integration provider limits the frequency of data syncs on lower plans. 15 minutes or Hourly are generally the lowest latency available on standard plans, with anything lower is reserved for Enterprise plans.

With Enterprise plans comes an upfront commitment. And this commitment is crucial to scaling cloud infrastructure as it allows the vendor to predict workloads, engineer scaling strategies and negotiate lower unit pricing.

Cost driver: Deployment engineering

A somewhat less appreciated aspect of building out a global SaaS product is managing the many deployments in different regions and clouds. In each deployment, a vendor must manage security, performance, uptime and more. The difficulty of this project is clear from the amount of flexibility vendors provide to their customers.

For existing SaaS integration tools, regional deployments are clearly a big cost driver. As an Australian company, we see first hand the challenge of investing in smaller regions and many SaaS apps never make the leap. A cloud vendor’s pricing reflects the cost and complexity of deployment engineering and those with the largest scale can offer the most flexibility.

Despite this, even the most popular integration tools limit the selection of cloud provider and region to Enterprise and Business Critical plans. Other integration tools have simple opted to stay as primarily customer-hosted.

Hybrid Data Storage, an Enterprise feature offered by ELT and Reverse-ETL vendors, is the perfect example of a band-aid solution to passing data through third party infrastructure.

Native apps allow smaller players to break away from SaaS unit economics. Piggy-backing off an established player like Snowflake, native app vendors one-up SaaS vendors on deployment offerings, reliability and security at a lower underlying cost.

Cost driver: Capital

Finally, while we've focused on architectural benefits, possibly the biggest cost driver is the cost of capital. As Tristan pointed out in his blog, along with VC industry commentators like the All-In Podcast, investor sentiment has rapidly changed since the peak in 2022.

"At the peak, valuations reached 100x forward ARR. Investors said: estimate what your annual recurring revenue will be in 12 months, multiply that by 100, and that’s how we’ll value the company… [Today,] Private multiples are now more like 10-20x forward revenue, so dropping 80%+"

To be attractive in the market, SaaS companies must show continuing increases in revenue. One of the reasons for Snowflake's success is that it shows 130%+ net-revenue retention. So customers spend more each year as they move more workloads over.

For SaaS integration companies, not only must they cover direct cost of goods sold, their pricing model must deliver ever increasing revenue. Levers like usage-based pricing is therefore a must to deliver returns to shareholders.

How a native app reduces costs compared to SaaS

Omnata’s native app delivers the lowest total cost of bi-directional integration on Snowflake by reducing the duplicated undifferentiated underlying costs of every SaaS product. Not only do the savings stack up across cyber-security and cloud infrastructure features, native apps deliver a superior architecture by design.

Reduced cyber security risks by design

Native apps run inside the trusted boundary of the Snowflake platform and a customer’s account. What’s more, an app must be explicitly granted access to data so even if your account is exploited, the blast radius is very limited.

Instead of paying multiple times for SSO, hybrid storage architectures or advanced governance features. Native apps like Omnata sit behind the platform features and frameworks that already exist.

There are multiple cyber-security benefits of Omnata’s integration native app;

• Credentials are stored in an already trusted platform

• Data is not passed through an external service

• A full history of data and logs can be retained

A native app vendor does not need to invest significant engineering securing its infrastructure and can pass on those savings to customers.

Reduced cost of securing standalone SaaS

Native apps enable customers to say goodbye to upsells to Enterprise plans for cyber-security features. This can save 50% or more of the license cost and reduce upfront commitment levels.

For customers, cutting out a SaaS app delivers related business savings:

• Easier approval that cuts weeks from buying cycles

• Eliminates privacy and residency issues

• Reduced compliance requirements

Snowflake Native Apps are fully compliant existing security framework and certifications.

Cut out duplicated underlying infrastructure

Native apps run in a customer’s account account using it’s compute.

At first you may think, that simply transfers cost from one stack to another, however native apps piggy back on Snowflake’s vast economies of scale. With SaaS apps, you duplicate the same costs between tools. Native apps utilise the same underlying infrastructure you already pay for.

In addition, ETL/ELT performance is also improved. External SaaS tools that poll Snowflake from the outside run less efficient diff queries resulting in excess credit burn and slow sync latency.

As a native app, Omnata’s Sync Engine retains a full history of data changes in the customer account enabling smarter incremental logic and more efficient queries. The increased Snowflake credit usage for the UI and background tasks is more than offset by the savings of not running a standalone SaaS app.

The most deployment flexibility

In the native app model, customers need only pay for infrastructure once.

Native apps are automatically deployed across regions and clouds without any of the engineering complexity that a SaaS company must take on. Running on Snowflake, a native app gains more deployment options than any integration SaaS tool on the market.

For a native app provider like Omnata, this reduces service costs significantly and these savings are passed onto customers in the form of flat connector-based pricing with no data volume charges.

Capital efficiency and marketplace distribution

Finally, native app providers can be far leaner than standalone SaaS vendors. Snowflake takes care of what would be a complex stack of technology.

• UI framework

• App discovery and distribution

• Billing engine and payments

• Logs and telemetry

• The list goes on…

Compared to SaaS companies that built during the peak of the MDS hype cycle. Native app vendors run a far less capital intensive business model, can build and ship new features faster, at a lower cost.

Try Omnata free for 30-days free

Get ahead of cost pressures and deliver more value from your data stack. Native integration is the first place to look to cut waste and deliver efficiencies.

Omnata Sync runs natively on Snowflake with:

• Direct connections to end-points, no data-handling infrastructure

• Bi-directional data syncs; ingest and output

• Faster cyber and procurement approval cycles

• No data volume charges

Visit the Snowflake Marketplace or get in touch with the Omnata team.