Big Data Grows Up at Strata+Hadoop World 2016

I recently spent time at Strata+Hadoop World 2016 in New York. I attended this event and its predecessor, Hadoop World, off and on for the past six years. This one in New York had a different feel from previous events including the most recent event in San Jose at the end of March. Perhaps because of its location in one of the financial and commercial hubs of the world, the event had much more of a business orientation. But it’s not just location. Past events have been held in New York also, and I see the business focus as a sign of the Hadoop market maturing.

Our research shows that big data can have significant business benefits. In our Big Data and Analytics benchmark research, more than three-quarters (78%) of participants indicated that predictive analytics is the most important area of big data  analytics for their organization. In our Predictive Analytics research almost three out of five (57%) organizations said they have achieved a competitive advantage through their application of advanced analytics. Thus we are moving beyond the early adopter phase of the technology adoption life cycle into the early majority. More and more organizations recognize that big data and advanced analytics can provide a competitive advantage. As a result, we see more focus on the business value of it, not just the technology required to pursue this advantage.

At the Strata+Hadoop World keynote presentations many vendors chose to bring their customers on stage or share stories about how their customers are positively impacting their organizations with big data technology. There were also plenty of technical training sessions, including two full days of training prior to the keynotes and expo, but the main stage of the event was focused on what you can do with big data rather than how to do it. The attendees also seemed to bring a business focus to the event. I spoke with multiple vendors in the expo hall who had attended both the Strata+Hadoop event in San Jose earlier this year and the New York event. They all described customer interactions that had more of a business focus than at previous events. People came looking for ways to apply big data technology to real business needs.

This is not say there wasn’t plenty of technology at the event including in particular data science, streaming data and data preparation and governance. Tutorials were offered on a variety of data science topics including how to implement machine learning in programming languages such as Python and Spark. Our research shows that Python is one of the most popular languages for data science analyses, in use by more than one-third (36%) of organizations. As I have written previously, Spark is growing in popularity as a way of providing big data, machine learning and real-time capabilities. At least half a dozen vendors ranging from large to small participated in the expo, touting their data science capabilities, and many other vendors’ marketing materials described how they support data science, for instance with data preparation tools that enable the data science process.

Processing streaming data in real time was also a frequent theme. Part of what makes big data big is that it is being generated constantly. It follows that you can probably get value out of analyzing that data in real time as it is being generated. In our research real-time analytics is the second-most frequently cited (by 54%) area of big data analytics, after predictive analytics. In its original incarnation, Hadoop was designed as a batch-oriented system, but as it has grown in popularity, much attention has been given to adding real-time capabilities to the Hadoop ecosystem, which I have described.

The themes of data preparation and governance come as no surprise. Our Big Data Integration benchmark research shows that reviewing data for quality and consistency issues (52%) and preparing data (46%) are cited as the two most time-consuming aspects of the big data integration process. Similarly our big data analytics research shows that data quality and information management is the second-most common barrier to big data analytics, cited by 39 percent of organizations. Vendors and the big data community are on the right track in addressing these issues.

The big data community continues to evolve, and the Strata+Hadoop World events are helping to foster dialog, education and growth. I’d say that this most recent event is evidence that the big data community is “growing up,” meaning that the focus has shifted to delivering business value. Strata+Hadoop World is a place where you can learn not only about the technology of big data but also how to solve business problems.

Regards,

David Menninger

SVP & Research Director

Follow Me on Twitter @dmenningerVR and Connect with me on LinkedIn.

Oracle Goes All In On Cloud

I recently attended Oracle OpenWorld for the first time in several years. The message at this year’s event was clear: Oracle is all in on the cloud. I had heard the message, but I didn’t get the full impact until I arrived at the Moscone Center in San Francisco. All signage at the event contained the word “cloud,” and Oracle issued 18 press releases in conjunction with OpenWorld related to cloud computing. I also found out that Oracle has its own definition of “cloud.”

Oracle now offers cloud services ranging from infrastructure as a service, which competes with Amazon Web Services, to database as a service to big data as a service to analytics as a service. These are in addition to Oracle applications offered in software-as-a-service configurations. Some years ago Larry Ellison expressed public resistance to “cloud computing”, but since then Oracle has been steadily investing in, adopting and now fully embracing it. Oracle’s direction reflects what our benchmark research has been showing for years: Cloud computing is being adopted ever more widely. For example, our Data and Analytics in the Cloud research shows that nearly half (48%) of organizations use cloud-based analytics today and virtually all (99%) expect to use cloud-based analytics eventually. The research also shows that one in four (24%) have the majority of their data in the cloud today and 86 percent expect the majority of their data to be in the cloud eventually.

In the big data and analytics market, Oracle offers the following cloud services:

• Oracle Exadata Cloud Service – massively parallel processing (MPP) SQL database
• Oracle Business Intelligence Cloud Service – business intelligence and visualization
• Oracle Big Data Cloud Service – Cloudera Enterprise (Hadoop) and data integration
• Oracle Data Visualization Cloud Service – self-service data visualization
• Oracle Big Data Preparation Cloud Service – self-service data preparation
• Oracle Big Data Discovery Cloud Service – data science with data preparation and visualization
• Oracle GoldenGate Cloud Service – data replication and streaming data
• Oracle NoSQL Database Cloud Service – key value store.

Oracle has also announced Oracle Essbase Cloud Service for multidimensional analysis and Oracle Big Data SQL Cloud Service for SQL on Hadoop and NoSQL. Oracle’s Big Data Compute Edition will allow organizations to scale Hadoop compute nodes and data nodes independently. All of these are indicated on the Oracle website as “coming soon.”

Faced with such a broad portfolio of big data and analytics capabilities, it may be a challenge for potential customers to understand the portfolio and decide which pieces are required for their organization. Fortunately, services based in the cloud are easier to try since no installation is required and subscription-based licensing doesn’t require long-term commitments to products.

Part of Oracle’s value proposition, based on its long devotion to the old model of on-premises licensing and management, is a mixture of cloud and on-premises deployments, often referred to as hybrid cloud. Oracle’s cloud services are available in three configurations: as a public cloud service subscription; as a managed private cloud service subscription managed by Oracle in the customer’s data center; and licensed as an on-premises deployment managed by the customer. Oracle is betting that this flexibility with be attractive to enterprises as they make their journey to the cloud. Amazon, the key cloud competitor highlighted in Oracle’s keynotes, does not offer on-premises or hybrid configurations. Our research finds that nearly half (47%) of organizations support integration of cloud-based data with on-premises data and 38 percent vice versa, suggesting a significant presence of hybrid deployments.

We should note that most of Oracle’s big data and analytic cloud services are not new. In fact, most of Oracle’s portfolio competes with other products that have been in the market for years. Its strength is to excel at making products enterprise-ready. Others may find new and innovative ways to tackle computing challenges, but as these innovations take root in the market, Oracle adopts them, hardens them and makes them available for critical applications. It also adds innovations around the edges, but fundamentally Oracle makes these capabilities industrial-strength for dealing with issues such as security, reliability, manageability and governance – necessities that are often overlooked as new products come to market. If your organization needs to support mission-critical big data and analytics, I recommend you consider Oracle’s offerings. They have the breadth and depth to meet most needs.

Regards,

David Menninger

SVP & Research Director

Follow Me on Twitter @dmenningerVR and Connect with me on LinkedIn.