"And the correct answer is...": providing effective decision support for your employees
Brett Stineman 270002944C email@example.com | | Tags:  healthcare decision-management banking business-events retail decision-automation watson business-rules decision-support predictive-analytics
0 Comments | 3,448 Visits
As we move forward into the new year, most organizations (including IBM) are implementing plans to improve their performance from last year. So, a good decision management topic to cover is how to help employees improve the effectiveness of their daily professional interactions -- with customers, partners and suppliers, as well as those for back office operations -- because every interaction involves some aspect of decision making. When added up, all those interactions (and decisions) either erode or elevate organizational effectiveness.
Why is effectiveness important? Effectiveness has to do with quality, or, put another way, the ability to "produce the desired effect" (Merriam-Webster, www.m-w.com). From a technology standpoint, many organizations are looking at how they can produce the desired effect, faster and more consistently, through better decision support to their employees. The use of a decision support solution can assist in this objective by narrowing down the number of potential options for a given situation and providing guidance in choosing between various options (e.g., using scoring to rank a set of viable choices).
In 2011, IBM received a lot of attention for its Watson system, which is able to analyze large amounts of unstructured information against natural language queries and come up with the highly accurate responses. While the initial demonstration of Watson was on the quiz TV show Jeopardy, a number of real-world uses are now being developed in areas such as healthcare; one example is focusing on providing disease diagnosis and treatment options decision support for medical professionals through its ability to analyze massive amounts of medical research. There are many other solutions where Watson can provide value with its unmatched power, but it's good to know that most everyday decision support scenarios can be handled through standard software applications.
Understanding decision support scenarios. In fact, a large number of decision support needs are for deterministic scenarios, where there is a known result for given situation. Returning to the example of healthcare, there are many cases where a critical patient situation can be avoided by checking against well defined restrictions or indicators. By running a patient's medical information through a rules-based application, medical practitioners can be warned if a certain medication could have an adverse effect based on factors such as age, weight, other prescribed medications, etc. The rules-based application acts as an automated assistant, reducing the risk for both patient and provider. As more patient information is collected digitally, event processing can be used in conjunction with business rules to alert care providers proactively, responding to specified trends in a patient's condition using readings from blood tests, EKGs, pacemakers and other medical devices. There are many deterministic scenarios in every industry -- the value of creating decision support applications in these areas is to ensure compliance and improve employee productivity by reducing the need to handle repetitive tasks that can be more efficiently handled through an automated system.
Of course, not all situations have a straightforward "given this, do that" type of definition. In many cases, there can be several appropriate alternatives or varying degrees of certainty based on the information that is provided. These are probabilistic scenarios, and this is an area in which various Watson-based solutions are being developed. But, here too, there are many examples of decision support applications being built using technologies such as business rules, event processing and predictive analytics. Using such common methods as scorecards or decision trees, which can be defined through data mining and analytic modeling then implemented as business rules, a set of options can be provided with guidance as to which have the best likelihood of success or preference.
For example, in consumer banking these technologies are being used to assist bank personnel in determining the appropriate products to offer a customer, with the added benefit of being able to validate eligibility for those products in real time. Retailers are adding decision support into point-of-sale systems to give their employees real time guidance around cross-sell and promotional offers. From a back office perspective, the combination of event processing, business rules and predictive analytics are being used in decision support areas such as risk/fraud alerts and case prioritizations in order to ensure people are focusing their attention where there is the greatest need.
So, how do you determine which approaches can meet the decision support needs for your organization?
By carefully considering each of these areas, you will be able to identify the best technologies for implementing a decision support solution that will enable your employees to work smarter and more effectively in 2012 (and beyond). To learn how TD Insurance uses a rules-based decision support solution to assist their agents, register here.