Page 3 of 3
Visual Analytics: BI is Dead, Long Live BI
Visual analytics technologies, which combine user-friendly data visualization and visual computation capabilities with sophisticated exploratory analytics, are taking the business intelligence (BI) industry by storm. Amid both industry hype and true momentum toward adoption, it looks as though visual analytics technologies are supplanting traditional BI.
Rather than be limited to carefully managed, IT-developed BI reporting and data access, visual analytics tool and application users are able to work in a more self-service fashion with less reliance on IT for their every dashboard, report, or other visual artifact. Users have more power and control over analytic discovery and can thus be more productive and creative with data, especially when trying to answer unanticipated business questions.
Yet, in survey research and conversations with business users and IT data managers for my new TDWI Best Practices Report, Visual Analytics for Making Smarter Decisions Faster, I found that in many cases, goals for visual analytics technology deployments line up with long-held BI objectives. Visual analytics technologies, particularly as they evolve toward more complete data preparation, greater comprehensiveness, and broader scalability, can potentially help organizations move closer to achieving BI goals and objectives.
However, if growth in the use of visual analytics tools and applications is not managed carefully, it could also exacerbate the very difficulties that enterprise BI rollouts were intended to solve. (Note: Please join the Webcast for this report on August 4; details are available at here.) Our research, as well as most industry research consistently finds that the "BI to the masses" goal of widespread, majority use of BI tools and applications is still just a dream in most organizations. Spreadsheet use is a major reason why.
Spreadsheet applications are commonly used for what we would call BI activities: data access, reporting, analysis, and presentation. Spreadsheets have long been the self-service data analysis application of choice for users, even in organizations where there is an enterprise BI standard.
IT's role in managing users' spreadsheet activities is typically small compared to the function's role for BI. Often, once users export data from general ledgers or other business applications to their spreadsheets, these become data silos -- "spreadmarts" -- that add to data chaos and make it difficult for IT to supply users with timely data.
Making Peace with Spreadsheets
BI and business-focused analytical application systems have helped organizations seeking to reign in spreadmarts chaos, in some cases by bringing spreadsheet use closer into the BI fold. For the report, I interviewed Greg Backhus, director of data warehousing and BI at Helzberg Diamonds, which has been implementing Information Builders' WebFOCUS.
"We have eliminated a lot of spreadsheets," he said, "but at the same time, we have embraced [Microsoft] Excel as an output format; we now know that users' spreadsheets are sourcing consistent, current data refreshed via WebFOCUS and based off of the data framework we want them to be based off of." Helzberg's dashboards offer users different levels of self-service applications that employ interactive tabs to provide users with a combination of ad hoc querying and subscription-based access to a library of parameterized reports.
Other organizations are looking to visual analytics tools and applications to help them move out of heavy spreadsheet use. I interviewed Kris Munson, director of pricing at Watts Water Technologies, where users had been relying on IT-developed reports and heavy use of spreadsheets for data analysis. Users have historically had to wait for IT to find system time to run reports, each of which provided only a portion of the data users needed. Munson has been able to allay IT fears about expansion in more independent user implementations of Tableau Software tools by showing that his group is reducing data confusion and actually improving the ability to track data lineage.
They are also improving sharing and collaboration: "Once users have something useful for the entire audience," Munson said, "we look it over and make sure everything is done right. We then put it in the production folder where everyone can start using it."
Business/IT Leadership for Data Discovery
Watts Water offers a good example of business users sharing responsibility with IT to make expansion of self-service visual analytics tool and application implementations successful. IT needs to play a critical role in ensuring satisfactory performance for queries against the data warehouse and other sources it manages as well as in governance, including data quality and security.
The best course is to pursue "managed" or "governed" self-service that features good communication and division of responsibilities. A best practice is to establish a governance or center of excellence committee to support growth in self-service visual analytics. In this way, enabling "the masses" to be more data-driven in their daily strategic and operational decision making can be a realistic goal.
Data-Driven Operations: Critical Focus
Our research over several years has shown that for most organizations, improving operational efficiency and performance through better information flow has been a key BI objective. Performance management dashboards and scorecards have helped improve communication of strategic goals to operational managers and users, who are often now held accountable for metrics associated with cost management, customer satisfaction, and more.
Unfortunately, BI tools have often proven too difficult for nontechnical managers and frontline users to implement as they try to analyze data tied to performance metrics. Easier-to-use visual analytics tools can also help users reach out to other types of data that might be beyond the narrow focus of performance metrics but that provides important context.
Swisscom, for example, is implementing SAP's Lumira to analyze a mixture of internal and external data drawn from third-party business partners and suppliers to support key performance indicators (KPIs). Matthias Mohler, BI consultant with the SAP consulting group at Swisscom, told me how users are able to interact with data more effectively through visual analytics, especially when compared to previous attempts at tracking KPIs with spreadsheets. Users are able to enrich data on their own, but most data preparation is still performed by experienced data analysts and managers. However, "users can now share and collaborate with data more than they could before," Mohler said.
Visual Analytics: To Boldly Go ...
To borrow the famous Star Trek phrase, rather than undo BI, use of visual analytics tools and applications could seek out new BI life where no one has gone before -- but not necessarily where many BI architects and dreamers envisioned the mission continuing. Users are finally getting tools that offer a brighter alternative to spreadsheets but that are not more difficult to use.
The challenge will be to deploy the tools and applications in a managed fashion that furthers the progress made previously with BI. As organizations push forward into new data environments, they must ensure that they do not erase the progress they have made with BI thus far.
David Stodder is senior director of TDWI Research for business intelligence. He focuses on providing research-based insights and best practices for organizations implementing BI, analytics, data discovery, data visualization, performance management, and related technologies and methods and has been a thought leader in the field for over two decades. Previously, he headed up his own independent firm and served as vice president and research director with Ventana Research. He was the founding chief editor of Intelligent Enterprise where he also served as editorial director for nine years. You can reach him by email (firstname.lastname@example.org), on Twitter (twitter.com/dbstodder), and on LinkedIn (linkedin.com/in/davidstodder).