Faster Performance and a 2x Reduction in Snowflake Cost

The Cube x Security Scorecard user story.

Security Scorecard logo
IndustrySaaS
Employees550
HQNew York, NY, United States
StackSnowflake, Observable
Use Case Embedded Analytics

Security Scorecards Launches Embedded Analytics

Security Scorecard’s mission is to make the world a safer place by transforming the way organizations understand, mitigate, and communicate cybersecurity risks to their boards, employees, and vendors. Today, thousands of organizations leverage Security Scorecard’s patented rating technology for self-monitoring, third-party risk management, board reporting, and cyber insurance underwriting.

In order to achieve their mission and continue to serve their customers, Security Scorecard constantly invests in new functionality that improves the customer experience. One such project was focused on analytics - reporting and dashboard capabilities - that they could offer customers within the product.

Evaluating and Delivering Embedded Analytics

Benja, VP of Engineering, knew that it was of utmost importance to deliver an embedded analytics solution that could handle their huge datasets without sacrificing speed - and to do it quickly.

To start, the team evaluated the embedded analytics capabilities of both Sisense and Looker. Frustrated with the lack of transparency of the Looker team and impressed by both Sisense’s demo and the attentiveness of their support team, Security Scorecard moved quickly to sign with Sisense.

After the contract was signed and they began work, however, it quickly became clear that Sisense was not going to provide what they needed. The Sisense team was very hands-on and helpful, but nevertheless, there were issues that surfaced that couldn’t be resolved. In addition, the developer experience was clunky and made it difficult to work efficiently. The team also heard from another Sisense embedded analytics customer that they needed a team of 10 engineers working on Sisense directly. Security Scorecard didn’t want to spare that many resources just to make it work.

Discovering Cube Cloud and Quickly Getting to Work

As he was considering what to do with their Sisense deployment, Benja heard about Cube Cloud and within a few short weeks was able to do a proof of concept and get hands-on with the product. He experienced how easy it was to onboard, a short learning curve, as well as a much more robust developer experience. Having tried to fit Sisense in for their embedded analytics solution first, Benja appreciated the “headless BI” architecture that allowed them to decouple their semantic layer from their visualization layer. Observable was chosen as their visualization layer and they were ready to move forward.

Marcus, Engineering Manager, took over the project and appreciated how easy it was for him to ramp up and understand Cube Cloud.

“We were under a tight deadline to get off Sisense. But I was able to pick up what had been done already just by looking – without having to read any documentation. The Cube UI is very intuitive.”

It took just 2 months to implement all six of our dashboards with Cube Cloud and Observable and the team has been using Cube Cloud now for a year and a half.

Cost-savings, Faster Development, Reducing Latency

When they were switching to Cube Cloud, they were told by other vendors that there would be a higher operational cost. However, that never materialized and in fact “the swap from Sisense reduced our total infrastructure cost by 60%” said Benja.

Once Cube Cloud was up and running, they realized how easy it was to release new dashboards and analytics. Before Cube, said Marcus, “It was easy to create a new data model, but working with actual data was a challenge. The team that made those endpoints available were always swamped with other tasks. Now with Cube, as long as the data is available in Snowflake, we can run queries ourselves. The real issue was that we didn’t have endpoints before and now, with Cube, we do. Cube unblocked that bottleneck for us.”

Over time, they found that “the cycle time for implementing new analytics and reports went from roughly 1 quarter to under 2 weeks in most cases” said Benja.

“We addressed multiple security concerns (caused by first trying embedded analytics with Sisense) given Cube's flexibility for custom access control, and the freedom to implement proper browser session security.” - Benja

In some cases, before Cube Cloud, the latency with embedded analytics was “so high that reports were just impossible to attempt” said Benja. With Cube, we reduced latency on those reports because of the flexibility provided by the headless BI architecture. Lately, the team has been working on adding more pre-aggregations. Marcus shared the “Gradual systematic work that we do with the Cube team to find the queries we want to optimize. Then we pre-aggregate and see the results.”

“It was a really big improvement. We reached faster performance for some queries. Even better, pre-aggregations reduced the number of requests to Snowflake by 2x which translated into a big savings on our Snowflake consumption.”

Overall, Security Scorecard continues to benefit from Cube Cloud in improving the performance of queries, reducing latency, and ease of development. Cube also delivers value in being agnostic to data source and visualization layer. Security Scorecard is considering moving from Observable to building their own custom visualization with React or similar libraries. They are also considering whether Snowflake is the right choice and trying out different cloud data warehouses. Because of Cube Cloud, they can easily move different tools in and out of their data stack - both on the visualization layer and the data source layer - while holding Cube as the underlying unifying semantic layer.

If you want to learn more about how Cube Cloud can deliver a semantic layer for all your data needs, please visit: www.cube.dev/contact.

Ready to upgrade your data stack?

Related Use Cases

Check out Cube’s other solutions

Related Blog Posts

Stay up-to-date with the latest from Cube