Business Survival Skill II: Making Decisions Leveraging Speed & Prediction

In the past, organizations could survive making slow and deliberate decisions. Today is a different era where the speed, accuracy, scope and completeness of a decision can make a difference in immediate profitability and longevity. The trend toward real-time analytics is transforming decisions in three ways. The first is the movement from pure historical data(what happened) to more in-flight and hypothesis focused data. The second is to use more streaming data that occurs over a tighter time frame. The third is that the analytics are being embedded closer to the real operations of an organizations; particularly in processes that run the organization. This is the second in a series. See

http://jimsinur.blogspot.com/2014/02/business-survival-skill-i-recognizing.html

Business Decisions:

History, Real Time and Prediction:

Organizations have to stop driving through the rear view mirrors. While history guides good decisions, better decisions can be made considering the most up to date information available and predictions based analysis. The combination of all three gives the business professional the best odds of making that educated decision that will drive for more success.

Years, Months, Weeks, Days, Hours, Minutes, Seconds & Sub-seconds:

Certain decisions have longer periods to brew, but one thing is for certain. If it took weeks in the past, it will likely require Days or Hours in our new world of high speed. In fact the premium on making operational decisions in real time is high. This is the new trend that is be accelerated by wider process and decision scopes and more sensors that are feeding decision opportunities faster than ever.

Strategy, Tactics and Operations:

It quite clear that operational decisions need more speed and strategy decisions need more prediction, but there is change to put strategy and tactics in business operations to make them more intelligent. This muddies the water as to what kind of decisions need speed and predictions. In fact there is a shift from mostly on demand analytics to the inclusion of more inline analytics. These inline analytics may allow for the adjustment of tactics or even strategies where scenarios and associated responses have been planned out and responses have been prepared.  


Technology Assisting Decisions:


Big Data in Memory:

The ability to gather and hold simple events is greatly enhanced by big data. For organizations that want a big view and high speed, big data in memory is a must. In-line analytics can be leveraged across large many events over time more easily. Patterns detected sensed in big data can trigger on demand analytics as well.

Real Time Activity Monitoring:

Activity that is unusual can trigger real time multiple analytic assessments (poly-analytics) as well as notify business process managers of a need to run on demand analysis or predictions to look for a change in tactics or strategy.

Process & Exception Management:

Quite often the early signs of an emerging pattern occurs in the exception grates of processes. It is easy to assign exceptions to skilled knowledge workers to complete case work, process  management that highlights and points out unusual exceptions is a growing trend thus triggering no demand analytics for changes in operations.

Assessment Agents:

There can be specialized agents that can be triggered in the event of an on demand need. An example might be an assessment agent that have been built for expected conditions. Also agents can operating in-line to notify business professionals of any need to adjust operations, tactics or strategy. 

Rule Engines:

Decision management software services provide prescriptive advice. They run on demand when a person, an application program or some other agent needs computational support for making a decision. For example, an application program may invoke a rule engine to score a customer's creditworthiness when he or she submits an order for goods. Business policies, goals and constraints are represented in the rules and algorithms embedded in decision management tools. In this example, credit scoring and other aspects of the decision could have been implemented in standard application code. However, using a rule engine simplifies application development and maintenance for complicated or volatile business policies. In some cases, business analysts or power users can directly modify the rules without programming, and with little or no IT staff involvement. 

Scoring and Optimization:

There are many scoring and analytic capabilities that can be leveraged. Examples range from data visualization, text analytics, forecasting, statistical analysis to contextual analytic methods. There is not end of combinations that can add value to speedier decisions that can help create better outcomes. 


Net; Net:

Fast and predictive decisions are dependent on situation awareness and predictions. Situation awareness is provided by systems that provide continuous intelligence – constantly updated descriptive analytics (although they also have elements of predictive analytics  because they may project what is likely to happen based on leading indicators and patterns of events that have been detected). The decide phase is when decision management techniques are applied to determine the best available solution (this is where prescriptive analytics apply).  More of these decision capabilities will be built closer to the business activity inside and along side processes that orchestrate people and systems