Putting Data to Work: Winning Approaches to BI, Analytics and Reporting

<< back Page 2 of 4 next >>

Bookmark and Share

A benefit to organizations is that the cloud movement is also enabling a tighter integration of analytics into the core business workflow, according to Aaron Rosenbaum, vice president of product strategy at MarkLogic. The cloud analytics momentum is being fueled by an increase in competitive pressure, which is driving customers away from reporting silos, as well as an appreciation for the cloud’s cost and scale abilities. The use of cloud for BI and analytics, however, may depend on where you are in the business. Cloud-based applications may be best suited for edge-of-the-enterprise requirements as they arise. “For lines of business, it’s easier to go SaaS, and embedded BI often comes with many cloud applications,” said Sumit Sarkar, chief data evangelist for Progress. “For organization-wide BI, on-premises still seems to be the best fit, based on existing skills and the     costs involved with moving to cloud.”

Khan cautions that some heavily regulated industries may be hesitant about positioning analytics data in the cloud, at least for now. “In industries such as health- care, with HIPAA, or financial services, with PII regulation, there is still a preference for on-premises to preserve data security and customer privacy. Private or dedicated cloud environments offer some flexibility to work around these constraints, but then organizations don’t fully get the benefits of SaaS models.”

Manish Gupta, CMO of Redis Labs, also issued some caveats regarding the rush to run analytics in the cloud. “Issues related to data size and location, network- and workload-induced latencies, governance and regulatory limitations, security perceptions, and enterprise infrastructure and process readiness will remain a hindrance to an all-cloud BI and analytics world,” he cautioned, observing that the future will be hybrid.

Data Democracy at Last?

With growing ease of use and cloud-based access to BI and analytics, there’s been a surge of non-technical users over the last 18 months, said Villacís. The increased data democratization brings with it a flurry of “visualization, mashups, and self-service data preparation,” he added. “People in different roles or functional areas are no longer just consumers of information but are now both analysts and consumers at the same time.”

The opening up of analytics to the enterprise is a relatively recent initiative. BI and analytics tools and platforms have typically been limited to handfuls of analysts and decision makers. The biggest change in analytics is IT’s move toward self-service while also addressing users concerns over lack of trust in the quality and accuracy of enterprise data, said Jean-Michel Franco, director of data preparation, data quality, and MDM at Talend. “Currently, most enterprise data stores are only accessible by a few, data-savvy elite, like data scientists,” he continued. “But IT is increasingly trying to sort out the best way to empower the people who are closest to the data—the business analysts or line-of-business managers—with direct access to enterprise data stores so that they can become more data driven.”

Increasingly as well, analytics capabilities are being embedded within applications and functions across enterprises, to the point where they are almost invisible. Sarkar reported he is “starting to see predictive analytics getting packaged into regular usage of applications.” For example, capabilities such as lead scoring may run through complex algorithms or machine learning and users might simply see a score on their screen to make a decision.

There are certain types of applications better suited than others to pervasive analytics. There will be more adoption of embedded BI and analytics capabilities that are specific to data and the subject matter of the core operational/functional applications, enabling decisions that are repeatable in nature, said Patel.

BI and analytics “as a specialized functional area is becoming less common,” said Rosenbaum. “For example, risk analytics used to be a major user of traditional BI and analytics. As operational systems consolidate, there isn’t the need to make a silo just for the risk group. And, as their work moves from decisions based on reports to algorithm generation, it’s not even relevant to run the analytics on a stale copy of the data. As a result, BI and analytics is becoming pervasive and distributed.”

BI and analytics solutions are no longer solely for data scientists, said Duckworth. Newer platforms have made it much easier for business users to become active users of BI and analytics, he noted. “This is being driven by the emergence of the citizen data scientist—those who aren’t true data scientists but who do more sophisticated analytics in order to do their jobs better, whether they are in IT or marketing or sales or operations.

<< back Page 2 of 4 next >>