An enhanced predictive analytics performance management methodology is provided in the Integrated Enterprise Excellence (IEE) Business Management System.
The predictive analytics performance management metric examples below illustrate how organizations can benefit from using predictive analytics to create insightful dashboards. The described business analytics examples provide direction on how to address the organizational need that was highlighted in the Forbes article ″Why Predictive Analytics Is A Game-Changer.″
This Forbes predictive analytics article stated: ″ Thus predictive analytics is emerging as a game-changer. Instead of looking backward to analyze “what happened?” predictive analytics help executives answer “What’s next?” and “What should we do about it?″
Enhanced Predictive Analytics Performance Management Example that Transitions from a Traditional Scorecard Format to an IEE Report: Traditional SAIDI Reporting
The data for the highlighted illustration below is from a utility company; however, the described situation is similar to many other business organizational metrics. More predictive business performance analytics examples will later be provided.
The particular metric dashboard that will be assessed (and used with permission) is: The System Average Interruption Duration Index (SAIDI). SAIDI is an electric power utilities reliability indicator.
SAIDI is the average outage duration for each customer served. Wikipedia notes that the Institute of Electrical and Electronic Engineers (IEEE) Standard 1366-1998 estimates that the median SAIDI annual value for North American utilities is approximately 1.50 hours (i.e., 90 minutes).
For one organization, the following SAIDI report-out is reported to executives as part of a PDF document.
This type of report-out is appealing for several reasons but also has issues:
- With annual objectives, this type of predictive analysis report-out initially seems to have merit; however, this metric is the result of a process. Processes don’t typically magically change from December 31 to January 1 of the next year. Also, the ending process overall response result for the year is not known until the last day of the year. For example, what does this predictive analysis indicates how well has the organization done relative to SAIDI annual goal throughout the year? This form of reporting does not really provide this type of information. It would be more desirable to continuously track the process monthly. If there are many years of stable performance, the average of all the data would be a good estimate for how the current process is performing. This value could then be assessed relative to an annual objective.
- With annual objectives, one might expect that certain times of the year that there is an increase in the probability of having an issue. One could view that this is a source of common-cause process variability and should not be reacted to as though it were special cause events. For a month-to-month tracked stable process, one could use predictive analysis testing of various hypotheses for significance; e.g., difference between time of year or regions. When business process analytics hypotheses like this are found significant, this insight can lead to what might be done differently to improve the process overall.
Predictive Analytics Performance Management Examples: Predictive SAIDI Reporting
Organizations gain much when they report process output data using as a 30,000-foot-level performance metric format. With this predictive analysis approach, the stability of the process is first assessed without any calendar boundaries. If the process is considered stable, a prediction statement can then be made. If the prediction statement is not desirable, the process needs to be changed to transition the process to an enhanced level of performance.
A 30,000-foot-level report-out of the SAIDI data used to compile the previous graphic is:
From this predictive analysis report-out, one quickly notes from the left graph that for over four years that the process has been stable. There have been no improvements or degradations in performance over this period of time. Some months of the year may typically have a higher response than other months; however, this would be a source of common-cause input variability to an overall annual rate.
For this business process analytics assessment, if the estimated annual mean rate of 5.83 is considered satisfactory, then no change is needed. A control mechanism for the process may need to be implemented to ensure that the process was continually executed like it had been done in the past. The 30,000-foot-level chart predictive analysis report-out would still offer a high-level tracking assessment of whether the process response has changed over time, although this feedback would not be timely.
One might have noted that the above 30,000-foot-level chart had data transformations. The details of when and how these mechanics are executed are only important to the person who is constructing the charts. Much documentation and training is available for these practitioners to learn these predictive analysis skills. However, the interpretation of the chart is quite simple for anyone who one reads the statements at the bottom of 30,000-foot-level charts.
Predictive Business Performance Analytics Examples: Table of Numbers, Stoplight Scorecards, Time-series Plot
Additional conversion examples from traditional dashboard reporting to 30,000-foot-level predictive performance metrics are:
- Conversion of an executive table-of-numbers dashboard to a predictive business performance analytics dashboard.
- Conversion of red-yellow-green dashboard reporting illustration from this one of the several noted predictive business performance analytics examples.
- Conversion of time series tracking illustration from these predictive business performance analytics examples.
- Conversion of wastage percentage table of numbers to business predictive analytics; i.e., business intelligence analytics.
- Conversion of on-time delivery performance non-actionable graphic to a business process analytics output that provided much process-change insight.
- Conversion of product service percentage graph that led to firefighting with another application shown from these predictive analysis examples
- Conversion of a table of numbers for tracking overall equipment effectiveness to a metric that can have timely updates through predictive analytics software
A summary of dashboard conversions to 30,000-foot-level predictive performance reporting illustrations are available in the article Transitioning Traditional Dashboards to Predictive 30,000-foot-level Metric Reporting Examples.
Applying Predictive Analytics Performance Management
The above described 30,000-foot-level business analytics reporting is applicable to a variety of situations, including for the use of predictive analytics in healthcare. This key performance measures analytics reporting approach can be implemented in an organization as part of its Integrated Enterprise Excellence (IEE) value chain. This effort can lead to improvement efforts that benefit the enterprise as a whole.
IEE addresses the business scorecard and improvement issues described in a 1-minute video:
A description of the IEE system is provided in the article “Positive Metrics, Poor Business Performance: How Does this Happen?“:
Books that describe implementation details of the methodology:
- 30,000-foot-level predictive analytics reporting Chapters 12 and 13 of Integrated Enterprise Excellence Volume III
- Integration of 30,000-foot-level predictive analytics within Integrated Enterprise Excellence Volume II
- Integrated Enterprise Excellence overview
- Business Process Management
Training in 30,000-foot-level reporting
Contact Us to set up a time to discuss with Forrest Breyfogle how your organization might gain much from an Integrated Enterprise Excellence (IEE) Business Process Management System and its predictive analytics performance management.