Get your organization started moving down the path of business intelligence now. Don’t let perfection be your enemy. This was the message – loud and clear – at Business Intelligence and Analytics: Unleashing Data to Drive Quality & Financial Performance, a two-day conference put on by The Center for Business Innovation. The group obviously meant perfection in terms of applications for accessing information. This is healthcare. The data must be pristine.
On the morning of the second day of the conference, I asked the audience to share two things – what they learned on the first day that surprised them and what would they like to learn more about either on day two or at future conferences. Overwhelmingly, the answer came back that they were surprised to learn they were not alone in confronting organizational resistance to getting started on business intelligence initiatives. Consequently, what they wanted were ways to overcome this resistance in their respective organizations. With more than one hundred people to talk with and over thirty presentations from providers, payers, accreditation groups, technology vendors and business technology consultants, hopefully we provided answers to help.
To punctuate the point that early solutions are not always easy or particularly elegant, Dr. Raj Gopalan of the University of North Carolina Medical Center used an especially effective analogy. On one slide of his presentation, he used three pictures to illustrate the progression of ground transportation.
The first picture was a real horse. Up to the late 1800s, this was the accepted way of traveling efficiently. The second picture was a mechanical horse, which was developed to imitate the functions of a living horse. For a short period of time in the 1880s, it was thought that this was the logical progression in ground travel and the one that would most easily gain acceptance. As a matter of fact, in 1893 a man named Lewis Rygg patented a design for a robotic horse. You can find a picture of the design on the Oricom Technologies website .
The third picture was the automobile, which is now, of course, our preferred method of getting around. Dr. Gopalan’s point was simply that we cannot expect perfection right away. As a matter of fact, the present-day automobile is by no means perfect. One hundred years from now, people will marvel at how primitive it was compared with what they are driving.
Later that day, three of us (Raj, Jason Oliviera from Kurt Salmon Associates and I) were finalizing the details of a workshop we were jointly facilitating the next day. I brought up the mechanical horse concept. Jason suggested that we cannot reasonably expect business intelligence solutions to arrive in completely perfect form. We have to expect to go through various stages of maturity, in much the same way the automobile went through hundreds or even thousands of refinements over the decades. After all, Jason argued, we didn’t just drop the horse one day and get behind the wheel of a car with rack-and-pinion steering, electronic fuel injection and a dozen cup holders. The early days of automobiles involved metal tires, wooden platforms and greasy engines that belched black smoke all over the passengers.
This combination of ideas got me to thinking about the challenges people are facing as they attempt to get attention and acceptance for their business intelligence initiatives. What happened with the horse-to-car transition? First of all, people had to let go of the notion that acceptable means of land travel had to involve muscle, bone, hooves and fetlocks. Furthermore, they had to let go of the idea that power had to look or even function like a horse. They had to jump to a different track, so to speak. Innovation in land travel for individuals involved wheels, brakes and internal combustion. In addition, it started in a primitive state. Through experimentation, usage and user input, it was continuously refined.
What about business intelligence initiatives? We have primitive reports and queries today. Plus, we are rapidly progressing in maturity with dynamic reporting, data mining, exception reporting, dashboards and scorecards. We will get to the point (sooner than you may think) of using predictive analytics, healthcare business intelligence search capabilities, text and unstructured data metrics, and we will see business intelligence embedded into larger processes and also shared across organizations, partners, customers and suppliers.
I often hear people on both the business and the technical sides lament that we are not using all of the power our data can provide. I agree. But instead of kicking what we have today, we need to honor it as the first steps on a longer journey. Plus, what we in the healthcare business intelligence field are building today sure beats patient records in three-ring binders, on 3×5 cards or, worse, not to be found at all.
Getting Things Going in Your Organization
Wayne Eckerson of The Data Warehousing Institute (TDWI) published a wonderful white paper that detailed the concept of a business intelligence maturity spectrum. Eckerson classifies business intelligence efforts in six stages that are paraphrased as follows:
Prenatal. Static reporting primarily coming out of financial systems, used to support operational processes.
Infant. Spreadsheets that come out of executive systems, used to inform executives on the state of the business.
Child. Ad hoc and drilldown reporting using data out of analytical systems to give managers and workers tools for making decisions.
Teenager. Dashboards and scorecards that access data from a dedicated performance monitoring system, used to align the entire enterprise along strategic objectives.
Adult. Predictive analytics built on data from the preceding stages to drive the performance of the business.
Sage. Business intelligence analytics embedded inside other processes throughout the organization, as well as outside the organization (customers, partners, suppliers), which are used to not only drive the business, but also to change the game by driving the market.
No matter where your organization is on the business intelligence maturity spectrum, it is essential to get out and bang the drum for using analytics better. And, it is essential to use language and examples that are in line with the key issues facing decision makers across your organization. Following is a list of key people as well as issues within their scope of management that can be positively impacted by business intelligence. It is important to talk with these people and their teams about these key issues.
Chief Medical Officer. Medical management issues, clinical performance improvement efforts, patient and provider satisfaction measures.
Chief Quality Officer. Quality accreditation reporting needs, quality improvement programs, pay-for-performance contract measures.
Chief Research Officer. Protocol and guideline effectiveness measurement, funding proposals for studies.
Chief Financial Officer. Revenue management, cost control, capital decisions.
Chief Operations Officer. Facilities planning and management, people and skills management, equipment and supplies management.
Chief Marketing Officer. Service line decisions, marketing message development, marketing vehicle effectiveness measurement.
Chief Information Officer. Analytical capabilities design, development and support issues, data management needs, data access requirements.
The point is that healthcare organizations need to move along the business intelligence maturity path and need to do it quickly to respond to the intense pressure from payers, purchasers, buying groups, governmental authorities and quality watchdog organizations.
No matter how sophisticated or primitive your organization’s business intelligence efforts are, get started and keep things moving. I recommend attending conferences to learn what other organizations are doing to promote analytical excellence. I also recommend intense reading on the subject and talking with people in your organization on the clinical, business and administrative side to learn how business intelligence can help them reach their goals.
Thanks for reading!