GoodData CEO Sees Growth in Analytics Biz Shakeup
Roman Stanek has been in the enterprise computing space for a while, starting with his 1997 launch of a Java IDE called NetBeans, which he sold to Sun Microsystems two years later. Stanek has been involved with a lot of IT endeavors since then, including his current company, GoodData, which develops analytics software for OEMs. But Stanek has grown a little bit impatient with the pace of technological progress in the BI space, and so he’s doing something about it.
“If you look at the BI market today, it’s one generation behind what’s going on in the cloud,” Stanek says. “It’s very monolithic. It’s very desktop oriented. There’s a mismatch between the volume of data in the cloud data warehouses like Snowflake” and what these tools can handle.
It’s not just the heaviness of desktop BI software, Stanek says, but also the movement of data en masse, first from the data sources to the centralized warehouses through ETL data pipelines, and then database extracts back down to desktops for analysis on a distributed array of desktops. All that data movement requires too much manual intervention, which invariably leads to divergent data definitions and different answers.
And don’t get Stanek started on the licensing approaches used by his bigger BI competitors, which he decries as onerous and self-serving. “The biggest hypocrisy of the whole space is they talk about data democratization and they charge thousands of dollars per seat,” he says. “If I want someone to see one pie chart of Tableau, I need to buy them a license of Tableau.”
With the launch of GoodData Cloud Native this week, GoodData is taking a major step towards providing the type of analytics experience that Stanek thinks users should have in 2021.
The new analytics offering ships as a self-contained Docker container that runs atop any Kubernetes environment, including on-prem and in the cloud. The software is designed to integrates directly with cloud data warehouses, including Snowflake, Amazon Redshift from AWS, Microsoft Azure Synapse, and Google Cloud BigQuery.
When a user manipulates data in the browser-based GoodData.CN interface, the SQL workload that it generates is executed in the cloud data warehouse, and the results are pushed down to the browser. That’s the best way to leverage the massive cloud resources that are available, Stanek says.
Alternatively, if a customer doesn’t want or need the GoodData UI, they can license the community edition of GoodData.CN for free, and get the Postgres-based SQL generation and execution component that can be called via APIs, without paying Stanek a dime.
The shift toward cloud-native executables and workflows is obviously important to GoodData. This is how modern BI and analytics tools should be developed today, Stanek says. “We believe the whole analytics, visualization, dashboarding–everything will have to be rebuilt and be much more cloud native,” he says.
But equally as interesting about the GoodData.CN offering and the data as a service (DaaS) experience that Stanek hopes to kickstart is that it’s not radically different from what GoodData has been selling for years, at least from the end-user perspective.
In fact, GoodData.CN utilizes the same SQL-generation engine and visualization components that the company has been selling to OEM providers. The biggest difference that end-users will notice is that it features the GoodData name, as opposed to the OEM company that slapped its own name on the white-label software.
Stanek estimates that more than 1 million GoodData seats have shipped over the years, although he says the actual number is impossible to know. The experience of being a pure-play OEM (until now) was actually beneficial to building the type of streamlined, cloud-based user experience that customers demand today, Stanek says.
“I joke about it that, in the first 10 years of GoodData, we have three versions running, two in the U.S. and one in Europe,” Stanek says. “It actually makes us nimble because we don’t have to worry about legacy. We don’t have a million copies out there that we have to go restructure.”
In fact, he says that being ensconced in the OEM world and outside of the traditional BI space this whole time was actually a blessing. “It gave us almost an unfair advantage in being able to see how to scale analytics, being able to solve some of these global problems, like how do you scale analytics to half million users without actually giving everyone a copy of the tool?” Stanek says.
Stanek predicts that more changes are in store for the BI and analytics space, which he says has been “stale for the last 10 years.” Massive consolidation will occur in the cloud, he says, to the point where there are just five or six massive vertically integrated cloud data warehouses that hold much of the data. ETL jobs that used to be performed outside the database are transforming into ELT jobs that are controlled by tools like FiveTran and Fishtown Analytics data build tool (dbt). The visualization interface is next.
“Someone called it an iPhone moment of this industry. The whole data value chain will re-arrange,” Stanek says. “For us, this is absolutely the right time for us to go out. We don’t need to be hidden behind OEM. We built a wonderful piece of technology that anyone can install in any cloud and moved the analytics from desktop to the data warehouse and make it real time.”
Over time, GoodData plans to launch additional offerings, including versions of the dashboard interface and the SQL engine that can be acquired in the clouds, as well as a data marketplace. For now, Stanek is content to have his analytics software running natively in the cloud.