GoodData partners with Amazon on Redshift: When data goes cloud, cloud data goes into analytics
Big Data
GoodData, one of the key players in business intelligence and analytics, announced integration with Amazon Redshift data warehouse solution, as well as a partnership with Amazon. These are significant not just for GoodData, or Amazon, but also as pieces of the puzzle in a shifting landscape, namely business intelligence and analytics.
ZDNet connected with Zdenek Svoboda, GoodData co-founder and VP of Platform, and discussed everything from data connectors to semantic models and from business strategy to how Kubernetes enables predictable pricing.
Agile Business Intelligence and the cloud
GoodData has been around since 2007. For a rapidly developing domain like the one GoodData is in, this is a long time. Many players have gone boom and bust, changed hands and fortunes since then. Case in point, the recent acquisitions of Tableau by Salesforce, and Looker by Google. So let's start by taking the time machine for a quick ride back to 2013.
That was the year we introduced the term Agile Business Intelligence (BI), as part of a research report for Gigaom. A lot has changed since then, including Gigaom, and most of the vendors featured in that report. But not the key premises of Agile BI, or GoodData's execution on this.
GoodData was among the first to realize that all your data belongs to the cloud and acted upon this. How? By offering its software via the SaaS model, as well as connectors to ingest data from as many sources as possible, including, prominently, cloud storage and other SaaS solutions.
GoodData was among the first to get that data is moving to the cloud, and follow suit.
Another direction that GoodData was a trailblazer in was embedded analytics: Enabling customers to include GoodData-powered analytics in their solutions. These things may look like table stakes today, but being an early adopter back then, and executing consistently, is why GoodData is still around and well-positioned.
Today, as Svoboda explained, the move to the cloud among GoodData clients is massive. Svoboda cited research from IDC, according to which cloud is growing at a 20% yearly rate at the expense of the on-premise data centers that are shrinking 5% every year. So GoodData is following the data, and going cloud with Redshift, Amazon's Data Warehouse:
"We see a lot of customers and prospects who use [Redshift] as their system of record for their internal reporting and analytics. However, when they need to share analytics beyond the perimeter of their organization, to their customers or business partners, they must use a different architecture. GoodData is designed exactly for this use case scenario".
When data goes cloud, cloud data goes into analytics
The idea seems clear: Redshift is great but does not work ideally for all use cases. GoodData lets you extend its capabilities. Win-win, right? Let's consider data warehouse market trends. According to research shared by Svoboda, Snowflake is leading in terms of spending intentions, with AWS, Microsoft and Google following, and everyone else lagging.
As Svoboda put it, Amazon has the largest market share (> 50%), Microsoft Azure is the fastest-growing, and Google Cloud is investing to catch up -- see Looker acquisition. So it makes sense for GoodData to be partnering with Amazon, considering it already supports Snowflake and Google Big Query. Svoboda explained there is a whole range of joint activities with Amazon scheduled around Redshift integration.
How will the integration work? In a way, this is business as usual for GoodData. GoodData already has an extensive line of connectors, enabling it to ingest data from over 150 sources. So it's just adding one more source. Data integration is the unglamorous layer on which everything else may be built, and Svoboda emphasized the "everything else" part.
GoodData wants to act as the enablement layer for its clients to offer analytics in their own offering to their own clients.
Data in modern organizations are a core asset, coming from many different sources, including different software, cloud storage, data centers, and so on. GoodData positions its SaaS as the hub to connect customers, partners, franchises, and a distributed workforce through the use of analytics on that data.
Amazon
We spent some time discussing this with Svoboda, which emphasized the fact that for GoodData, enabling others to relay analytics-driven insights to their channels is what embedded analytics is all about. He went on to add that this represents about 50% of their use cases and 80% of their revenue, serving a network of something like 30.000 customers.
Another important part of GoodData's value proposition is what they call the Semantic Layer. Simply put, as the number of people sharing access to, and interaction with, underlying data grows, a common layer of abstraction and unification becomes necessary, to have a shared view of the data.
Breaking down a potentially vast landscape of concepts about data to sub-domains, for example, marketing or operations, ensures people working in the same field have a shared, manageable view of what they need to be aware of in their workspace. Then they can use this to define metrics and derive insights.
Coming up next: Kubernetes, definitely, and knowledge graphs, maybe
To us, this approach makes perfect sense. To derive insights, you need to integrate data not just on the technical level (dump everything in a data lake, for example), but most importantly, on the semantic one. Although the technology used for this by GoodData was not discussed, Svoboda agreed that conceptually, this sounds like a knowledge graph approach.
We don't know whether GoodData will be walking down the knowledge graph path for its semantic layer, though we do think it would make sense. What we do know is that adopting containers and Kubernetes, among the defining trends of the 2020s, too, is in GoodData's roadmap.
As Svoboda explained, GoodData will be going all-in on Kubernetes, offering its solution in a containerized version in the coming year. The reason, Svoboda said, is not so much to tick a box in the trends list, it's mostly because customers are asking for it, and it makes sense in today's landscape.
Getting all your data in one place is necessary, but not sufficient to get insights. To do this, a semantic layer is needed.
So GoodData will go beyond SaaS, enabling its customers to deploy and use it in any environment -- their own data center, or in any cloud provider. Welcome to the multi-cloud. But beyond usability and responding to customers, Svoboda argued, there is another reason for this: predictable pricing.
Leveraging the elasticity that containers can offer on the technical layer, via more dynamic management of resources, Svoboda argued, can go all the way up to pricing offered to customers. GoodData wants to predictably grow with each new customer's customer -- the "enablement" aspect.
This gives ISV customers better cost predictability if you can avoid charging for unpredictable CPU load, network bandwidth, etc. Containers and Kubernetes are the keys to achieving this, as per Svoboda. It may not sound fancy, and it sure will be hard, but the reasoning makes sense.
Stepping out from the time machine, back and forth in the past and the future, it seems to us the core of GoodData remains the same. Same as in 2013, GoodData in 2019 seems to be onto the defining trends of the time, and well on its way in execution. We expect to see them make their way in the 2020s, too.