Information blindness

Many leaders are crying out to see – to both access and visualize information that may be spread across multiple platforms and geographies. Here's how one healthcare organization healed themselves.

Are your leaders suffering from “information blindness”—a condition where information exists but cannot be seen in a format or timely enough manner to make sound operational decisions? Does it take too long for your Business Intelligence (BI) team to warehouse the information you need or update data tables at the speed of business?

These are common problems as business leaders become more dependent upon timely information to remain competitive, achieve operational efficiency and effectiveness, and make accurate daily decisions to keep the business on track. This problem is exacerbated as mergers and acquisitions introduce new entities to the enterprise that use different IT systems and therefore disparate and disintegrated information.

Many leaders are crying out to see; to both access and visualize information that may be spread across multiple platforms and geographies. One large rapidly growing health care organization was dramatically afflicted with this information malady. Here is their story and how they found healing.

Background

Vice presidents and department heads complained they did not have all the information they needed to run their departments, did not know where the information existed, nor did they know all the tools available to access the information.

To get a handle on the problem the project leaders contracted with a consultant to research, document, and graphically map all data sources and how the data flowed, as well as all the reporting tools available. Though internal staff could have done the job, they were already oversubscribed, so outside labor that could spend full time on the effort was a good solution.

Through this effort It soon became apparent the organization was data rich (the needed data existed), but information poor (accessing and making sense of the data was extraordinarily difficult). Across the health system’s different hospitals there were eight or more different electronic medical records (EMR) systems; four different labor productivity systems; three different financial reporting, payroll, benefits, accounts payable, and general ledger systems; and multiple supply chain, benchmarking, and revenue cycle systems.

Sometimes the data was even managed manually on spreadsheets. In technical terms, it was a mess. There was an active BI data warehouse project in place and making good progress, but currently  it only housed half the information needed to deliver the answers executives were seeking; and the ETA for building all the needed information into the warehouse was three years.  

Working with the consultant, the project team decided traditional Business Intelligence methods though powerful and proven, were not the best solution for this situation. They needed to come up with something different to meet the organization’s immediate needs without undermining the long-term BI Data Warehouse initiative. After a few ideation sessions, the team decided a Data Discovery initiative was the best cure for the organization’s information blindness.

Solution

Data Discovery is a relatively new approach to BI, coming into vogue around 2012. It is different from traditional BI methods such as Enterprise Data Warehouses (EDW), Ad-hoc reports, and Decision Support Systems (DSS). It was developed to reduce the time and resources needed to access data and to access it wherever it resides, including an existing data warehouse. The approach requires the customer to be directly involved in the development process. The process employs iterative prototyping to discover more precisely what the customer needs. Changes are made on the fly to the reporting application—usually a dashboard—and purposed tools present the information in a visual format that increases insights.

Sometimes customers don’t know if they really want the information they asked for until they see it modeled, and upon seeing it, usually have significantly deeper insights into the data they could not have had up front. The Data Discovery method leads to interactive refinements of the output until a better product is achieved than the first requirements would have delivered.  Traditional BI is heavily technology driven, whereas Data Discovery is user driven.

One executive complained that in the traditional BI approach he got tired of waiting three months to get the wrong answer. Why is the answer wrong? Because up front, the user does not always know the question to ask, and only after tradition BI tools source the data, extract, transfer, load, clean, and conform it is an answer delivered so the customer learns that she was asking the wrong question, or it was interpreted incorrectly by the programmer.

Data Discovery is user-centric, dynamic, and empowering—it builds the extract transfer load process directly into the discovery process, adjusting requirements on the fly.  Because rework is practically eliminated, and testing takes place interactively, a Data Discovery initiative has the potential to deliver executive insights in a fraction of the time of traditional BI.

The following graphic contrasts the traditional BI approach to the newer Data Discovery approach:

data discovery Avery Cloud

Outcome

Blindness cured. Employing the Data Discovery approach to five macro business processes, the team was able to deliver executive level dashboards in 90 days that were three years or more away waiting for the enterprise BI Data Warehouse solution.

During the project numerous unknown errors were found and corrected in the data; a solid data dictionary was built for each business process; users were trained to make their own modifications to the dashboards, freeing them from long waits for central BI resources.

In addition, the resulting dashboards required far fewer people to maintain, which is a true benefit as budgetary pressures make it increasingly difficult for BI leaders to employ a staff large enough to deliver the service levels demanded by the organization.

In each of the five business processes chosen, pleased operational owners proverbially said: I was blind but now I see.

This article is published as part of the IDG Contributor Network. Want to Join?

NEW! Download the Fall 2018 digital issue of CIO