In addition to opening up across enterprises, the power of analytics—once only available to large organizations—is increasingly now available to small and medium-sized businesses. “Today, the availability of rich data—including video data, low cost computing power, and analysis and visualization tools for the everyday business user—has made a more competitive playing field,” said Slaven Sljivar, vice president of analytics for SmartDrive, which offers performance insight and analysis to fleets and drivers.
Where’s the ROI?
However, there is work to be done, and some industry watchers argue that business leaders are waiting to see payback for the investments they have already made. “There has been a lot of hype around big data and self-service in the past 12 months, and, yet, when I speak to audiences at conferences, nearly 100% of end users state that they are not realizing a return on investment for BI and analytics,” said Rado Kotorov, global vice president and chief innovation officer at Information Builders, a provider of BI, analytics and integration solutions. “It seems that every 5 years the industry does more of the same thing—creating a new analytical tool that marginally improves usability and adoption for the analysts. This can be largely attributed to lack of widespread adoption within the enterprise.”
Like it or not, spreadsheet culture continues to dominate BI and analytical activities. “In the last year we still haven’t replaced Microsoft Excel with anything more ubiquitous and user-friendly,” said Diane Robinette, CEO of Incisive, which provides software solutions for spreadsheet analysis, control, and management. “While many BI solutions have been introduced to address Excel’s shortcomings, most traditional BI solutions come with their own set of usability issues.”
Still, while spreadsheets may be lingering, the era of staid reports is quickly receding in the eyes of other experts. “We’re seeing a real shift in thinking from BI and analytics as reports on warehoused data—either canned or interactive—and toward a more discovery-based model,” said David Gorbet, senior vice president of engineering for MarkLogic, provider of an enterprise NoSQL platform for big data applications.
The businesses reaping advantage from data analytics are those which have learned how to tear down the walls that separate information within their organizations. The trick to really getting value from big data is figuring out how to break down data silos, bring the data together, and then producing relevant, actionable information that has business benefits,” said Rachel Delacour, CEO and co-founder at SaaS BI vendor BIME. “This requires the query-blending technology that some BI solutions offer and that organizations need to adopt in order to quickly mix a variety of data sources.”
The most powerful solutions “are the ones that just remove the barriers between the time that a business person has a question and acquiring the data that can answer that question,” Alation’s Sangani added. “We’re finding progressive CIOs at modern digital companies such as eBay and Square alongside traditional grocers, auto manufacturers, pharmaceuticals, and banks doing the exact same thing —they’re investing in solutions that make the raw data available alongside the rules of how to use the data and shifting the control point from the production of the mart to the point of consumption.”
The cloud also is a potentially disruptive force in the BI and analytics space, but experts have varying views on how long the transition to cloud will take. “On-premises is a defunct old model—there is no more point in even highlighting the difference between on-premises and cloud,” said Bobby Koritala, chief product officer of Infogix, a provider of data integrity and analytics solutions. “All new relevant innovation is being driven through the cloud—in terms of technology and delivery.”
However, that view is not shared by all. Sangani noted that “most firms have hybrid on-premises and cloud analytic architectures.” Most of the customers his company works with have about 10% to 20% of their data sitting in cloud services, he added. “We’re finding that most customers still store most of their data in traditional relational marts, warehouses, and in on-premises implementations of Hadoop.”
Most BI and analytics spending “will still largely be on premises, even if it’s shifting,” Qlik’s Sommer agrees. “Very few companies are currently willing to run their systems of record data warehouse in the cloud.”
Some applications, by necessity, have to remain on-premises, at least for the foreseeable future. “Financial applications such as credit risk solutions often need to be run as on-premises solutions due to the nature of the data and regulatory requirements,” said Penny Herscher, CEO of FirstRain, a provider of business analytics. “Globally, there are also different requirements,” such as personal data about individuals that requires specific treatment with on-premises analytics, she added.
Many industry watchers also agree that the hybrid cloud approach will dominate the enterprise computing space for many years to come. “The reality is that many organizations are operating in a hybrid cloud world,” said Adam Kocoloski, distinguished engineer and CTO of IBM Cloud Data Services. “They see the agility and flexibility of the cloud but they also see the tremendous value in on-premises data processing. The key is tailoring analytics approaches that take advantage of both resources.”
Ultimately, cloud may offer approaches that suit the needs of analytic efforts as they scale and thus will increasingly become a more attractive option than continuing to invest in on-premises solutions. “The bottom line is that enterprises are looking for a data management infrastructure that provides them with maximum flexibility and agility to adapt to unanticipated future needs,” said MarkLogic’s Gorbet.