05 Nov John Purvis: Corda Technologies chief provides straight talk on the cultural realities of BI
Madison, Wis. – Wouldn’t it be transformative, when figuring ways to navigate though the current economic doldrums, to have business intelligence capabilities so advanced that just about anyone on your staff can drill down far enough to find data that can produce the right decisions?
That certainly would be a competitive advantage in a period of uncertainty, but the vast majority of businesses – even those that have invested in BI tools of Cognos, Business Objects, and others, haven’t reached that point. There is no doubt that “BI” is making a comeback, even enjoying a renaissance, but unless cultural barriers are overcome, the whole concept could once again wear out its welcome with broken promises.
In an interview with WTN, John Purvis, president and CEO of Corda Technologies, said much of the BI renaissance was due to an awareness of unfulfilled promise. In the past, the industry made too many promises that could not be kept due to an extraordinarily difficult set of business problems that vendors tried to address from a technology centric framework.
That was a big mistake that BI vendors cannot afford to repeat, he said.
“I think these failed promises were derived in large part from one, trying to solve a problem which had cultural aspects, with complex business interconnection aspects with pure play technology solutions,” he said. “Second, they were trying to offer a massive, enterprise level solution addressing all the important cases and therefore with all the attendant expense.”
Today, he said, the industry has a better understanding of the role BI can play in business performance management, and is trying to hit a “sweet spot” where complex analytics can be presented, in an understandable way, on a performance dashboard. As a result, Purvis does not believe vendors alone are creating a BI buzz for commercial purposes.
“I don’t believe it (the interest in BI) ever really went away, but I think there was a great deal of dissatisfaction with the failed promises of BI,” he said.
Corda Technologies has been in business for 13 years, starting as a charting and graphing tools company, including some patented IP, and later developing geographic mapping products. Corda, which has partnered with Wisconsin’s AE Business Solutions, got into business performance dashboard software after the balanced scorecard achieved a new level of interest. The company knew patented IP in the form of the data funnel, which allowed for heterogeneous access to data that could be nested and aggregated across an organization for the purpose of displaying it.
The most difficult part of BI, Purvis said, is presenting data in a useable form. “A lot of that is tied to the frustration of drowning in a sea of data, and not being able to do anything with it,” he noted.
For those who wonder how we can have so much data, have literally millions of dollars of BI implementation, and still wind up in a financial crisis that is partially due to a lack of internal controls and oversight, the answer is simple: financial institutions were drowning in data but we’re starved for knowledge. While BI was installed, analysts were only looking for certain parameters and probably ignoring data that may have uncovered the extent of the growing crisis. The right questions were not being asked, and no BI system is going to automatically start red-flagging things.
In BI, the challenge is that most of the tools require some degree of business analysis training for a broad array of users so that business analysts are not the only ones empowered to find an answer.
“You’re talking about alignment, not just being able to access data for queries that you can anticipate ahead of time, or for ad hoc queries that you oddly can anticipate ahead of time,” Purvis said. “You really are talking about using data to align people around the organization’s goals.”
Purvis said the most critical thing is the performance of the organization – business performance management – whether it’s a corporation or whether it’s a government agency. That holistic view of the business or agency has to take into account the balanced triangle of constituents where shareholders and customers and employees have equal interest in the ongoing success of the enterprise.
“I believe it’s possible to become obsessed with data, but until you have something that understands or is built on the aspiration of understanding what it means to the enterprise, and have a prayer of being able to pull it off in an excessively refinable way, I think you have nothing,” he said. “It implies a data democracy which frankly does not exist in most corporations.”
This is a significant obstacle for the promise of business performance management, and so is the protection of data silos among various business units that have no spirit of collaboration when it comes to ownership and control of data, Purvis agreed.
On the BI market as a whole, some of this attitude is promoted by industry consolidation such as IBM’s acquisition of Cognos, which has left the market with few pure BI players. Purvis believes this return to a more siloed environment in the industry reinforces cultural silos within organizations, and that’s why it’s more of a challenge than ever to democratize data and achieve pervasive access so that more people in an organization can make “rapid decisions on rapidly deployed data.”
Getting to the point of excessive refinement of BI and pervasive access to data is a challenge for any organization in any industry, but healthcare may have the longest path. In the healthcare space, business intelligence still is taking a back seat to the adoption of electronic medical records and other healthcare IT.
Two exceptions are Dean Heath Insurance and Group Health Cooperative of South Central Wisconsin, both of Madison. In each case, a key business driver is a high rating from the National Committee for Quality Assurance.
Dean Heath Insurance is moving rapidly on BI, going live in June with its new DeanInsight system, which also is being used by Dean Health Plan and soon by Dean Health Systems and SSM Healthcare.
Karl Richards, vice president of technology for Dean Health Insurance, said the insight product will drive analytics for Dean’s provider organizations to help them better manage the care of diabetics, asthmatics, and other patient populations.
DeanInsight is a combination of a Business Objects tool on the front end, and its use of extraction, transformation, and loading (ETL) tools from Oracle and Informatica that enable it to transfer data from various sources into a data warehouse. From there, Business Objects software allows Dean to build “some funky dashboards” for both print and web-based reports, Richards said.
Richards agrees with the premise that organizations can’t begin to conduct thorough analytics without an EMR, and noted that the Dean organization, when all of its hospitals and facilities have completed their EMR adoptions, will have the most extensive EMR roll out in the country.
“That will help us take BI a lot farther,” he stated.
At Dean, data already is democratized to some extent, but like everything else associated with technology and culture, it’s part of an evolutionary process. Just about everyone is expected to know how to pull out a corporate dashboard, and then drill down to dashboards related to their business functions. Access to various data portals is restricted by role-based security, not departmental turf battles.
For Group Health Cooperative of South Central Wisconsin, BI continues to be an area of emphasis. Galen Metz, information services director for GHC-SCW, notes the organization has adopted virtually all the core Epic Systems’ ambulatory and insurance modules, from the core electronic medical record to the member portal. It was the first Madison health organization to deploy an EMR, and it’s far enough along the adoption road to focus more on the business intelligence advantages of EMRs.
Metz continues to believe that business intelligence is the next technological wave in healthcare after EMR implementations because BI can leverage data to improve healthcare quality, costs, processes, and business decisions.
The organization has navigated several barriers to put itself in a position to leverage data, and other healthcare organizations will confront similar obstacles, Metz cautioned. They include the difficulty of focusing on BI when also trying to implement an EMR, the difficulty of building support from upper management (key for any large-scale project), and the need to secure funding for the initial step of developing a data warehouse. Efforts to load the data, often from disparate sources, present another challenge, as does data cleansing because often the data capture requires more than “just bringing everything together,” Metz explained in an e-mail.
According to Metz, the organization expects BI to be a key component of its 2009 strategic initiatives. For example, dashboards are an integral part of its process-improvement effort, he indicated.
In terms of data control, Metz characterized the model as a hybrid that centralizes the complex, analytical BI activity, and decentralizes access to and variations of existing reports. A centralized BI team creates new reports and provides analytical support.
“About 25 percent of our employees have access to a self-service product (Business Objects’ InfoView),” he explained. “This enables users to find existing reports or create ad hoc queries. For example, one could run an existing report with different parameters, like a different date range.”