SAP has unveiled a new service aimed at easing the ability to manage and use data by creating a data fabric architecture or a single environment that integrates data from sources across the organization.
Built on SAP Business Technology Platform (BTP), SAP Datasphere is an evolution of the existing SAP Data Warehouse Cloud and SAP Data Intelligence Cloud, and the service is available now as part of the BTP consumption-based cloud services, according to the company.
Enterprise data is scattered across external and internal systems, including on-premises applications, cloud hyperscalers and data vendors, and accessing and using that data is complex and challenging, said Irfan Khan, president and chief product officer at SAP HANA Database and Analytics.
In order to make data useful for decision-making, companies have had to move data from systems or applications to central storage locations, Khan said. In doing so, however, they often lose the business context around the data, which must be restored with a great deal of time and effort.
For example, to do analysis on supply chain disruption, customers have to copy supply chain data into an analytics tool to combine it with data from other sources, he said. However, critical metadata such as business partner connections is lost when the data is extracted and must be added back in manually or reconstructed, and the burden is on customers to do the work.
SAP Datasphere enables customers to unify disparate data sources by building a data fabric architecture that integrates data from SAP and non-SAP sources while keeping the business context and logic intact, Khan said.
SAP Datasphere is available now as a part of the SAP BTP cloud platform, with cost based on customer consumption.
Partnerships with data vendors
SAP Datasphere includes partnerships with the cloud data storage and management platforms Collibra, Confluent, Databricks and DataRobot. Each vendor brings a set of capabilities to SAP Datasphere for customers to use, according to Khan.
Collibra's capabilities include data governance, privacy and cataloging; Confluent is used for connecting and processing data streams; Databricks provides its Lakehouse data lake capabilities; and DataRobot has automated machine learning training and AI data preparation.
The partnerships will enable customers to use tools that they already have to build out a business data fabric, Khan said.
"We are taking on a true data ecosystem approach, not a go-to-market piece where we co-sell or they resell," he said.
Not new but addresses real problems
SAP Datasphere isn't new, but it does address data issues that face companies, according to analysts.
Stewart BondResearch vice president, IDC
"[SAP Datasphere] is a rebranding of existing capabilities, but what's interesting with this new branding is how SAP is approaching the problem, focusing on the concept of a business data fabric," said Stewart Bond, a research vice president at IDC.
The partnerships show that SAP is aware that it needs to expand the data landscape to be successful, he said.
"Clearly SAP realizes that not all the data is going to be in the SAP Datasphere and that the SAP Datasphere does not have all the functionality to manage all disparate data sources in the organization," Bond said.
SAP Datasphere addresses a fundamental problem companies face today in getting better value from their data, said Joshua Greenbaum, principal at Enterprise Applications Consulting. It also represents a much-needed continuing change in SAP's mindset by allowing the commingling of SAP and non-SAP data.
"Fundamentally, it's important to look at the outcome here, which is that you get to use data in a more strategic way and with better governance and better curation," Greenbaum said. "They are recognizing more and more that their customers need to deal with heterogeneous environments, so it's a positive move all around."
Moving data is the "third rail" of any transformation effort, he explained. It propels the effort forward but getting it wrong can kill the analytical side of the transformation.
"Just getting that curation side and governance side right is really important," Greenbaum said. "Being able to turn around and use that data to drive business outcomes is the icing on the cake, but it's very hard to do and companies really struggle with it."
Partnering with the data vendors represents that change of mindset, as SAP needs a thriving ecosystem going forward.
"They need these partnerships because it's going to be up to the partners to help get customers through the next mile," Greenbaum said. "These are not the usual 500-pound gorilla partners, these are smaller strategic partners, which is what SAP needs a lot more of -- and there should be more of that across the board."
However, SAP needs to be careful not to oversell the idea that integrating data at enterprise scale will be as easy or "seamless" -- as the marketing for Datasphere appears to indicate, said Jon Reed, co-founder of Diginomica, an enterprise software industry analysis firm.
"When you use words like seamless integration, you must be clear that it's aspirational," Reed said. "This is not objectionable as a concept, but it is as a description and your skepticism starts to mount when you hear that."
Reed said he also believes that tying the SAP data in with non-SAP data should assure customers that they're not necessarily tied to one way of massaging the data within an SAP context in order to get value.
"SAP already had APIs to access some of the non-SAP tools, but they're trying now to provide the business context," he said "Because a lot of times when you pull data from a system, you lose the metadata and the business context and you don't get the same value out of it. That's the big development in this release."
Customers should also keep in mind that SAP Datasphere is still in the formative stages and there will be more clarity of the value with more use cases, Reed said. There will also need to be more clarity around licensing questions.
"It's a bonus for the customers on the Data Warehouse Cloud because they get access to more features, but there will be licensing considerations for new customers," he said. "They'll have to explore this based on whatever tools and analytics they might have from other vendors."
Jim O'Donnell is a TechTarget senior news writer who covers ERP and other enterprise applications for TechTarget Editorial.