Advanced AnalyticsBusiness Intelligence

The Era of Decision Intelligence

Have you ever heard the expression, “Good decisions don’t just happen?” Whenever I hear that expression, I think about what goes into making a decision. As with any discipline, there is an art and science to making decisions. Some people may argue that the advent of artificial intelligence (AI) and machine learning (ML) is rendering the need for keen decision-making skills obsolete. These people would be incorrect. While AI and ML can crunch millions of data points at a speed no human can match, these technologies are still a reflection of a person’s representation of a mathematical model. Especially in complex, multi-variant problems, there is much more to come to an optimal, balanced decision.

Unfortunately, making decisions isn’t getting any easier. Today’s decision-makers face more challenges than ever before. Business cycles continue to shrink, adding time pressures to decisions. Massive quantities of data are produced in support of decisions but determining what data to use and when to use it presents other challenges. Finally, in a hyper-connected world, decisions can have far-reaching impacts. In many cases, impacts beyond the original intent of the decision-maker.

For these reasons, we see the rise of decision intelligence in many organizations. Gartner Group predicts by 2023, more than 33% of large organizations will practice decision intelligence, including decision modeling.

Decision intelligence is concerned with all aspects of selecting between options. At its core, decision intelligence turns information and insights into better actions. Decision intelligence invokes a wide range of decision-making methods to design, model, align, deploy, and monitor decision models and their associated processes. As such, decision intelligence brings together several disciplines, including decision management and decision support.

To leverage decision intelligence, an organization must define the problem they are seeking to solve and how they might go about solving it. For example, metrics and measures commonly reviewed, the level of data quality required (absolute or “good enough”), emotions or biases that impact the decision, consequences and ethical implications of the decision, and balancing the constraints to arrive at an optimal, balanced decision.

As decisions have become more complicated and more time-sensitive, decision-makers have become more reliant on higher degrees of automation and supporting analytics to add context and framing to a decision. However, decision-makers aren’t ready to totally abdicate their role in the process. Quite the opposite is actually the case. Since few decision-makers have a background in mathematical sciences, it is difficult for them to believe a “black box” can achieve the same or better outcomes than their experience and intuition. Not surprisingly, many leaders, when confronted with the prospect of losing control of decisions, will challenge or second-guess the results of advanced analytics or AI models.

Decision intelligence works to operationalize sustainable decision models. Decision models follow a set of structured steps to identify and present potential resolution options to a decision-maker. These steps include:

  • Observation – identifying the decision to be made and collecting all of the relevant data needed to support the decision
  • Assessment – using the collected data, perform a situational assessment and determine potential actions
  • Modeling – develop model-enabled scenarios that result in alternative actions and their expected outcomes
  • Contextualization – balancing the numerous constraints and objectives of the models to present the decision-maker with a set of potential deployable actions
  • Deployment – deploying or executing the selected action
  • Monitor – measuring the impact of the selected action and the potential impact of actions not selected

One particular model of decision intelligence that is seeing the digital assistant. The most prevalent and basic form of a digital assistant is the voice assistant. Siri, Cortana, Alexa, and Google are the most popular examples of a digital voice assistant. However, when it comes to complex, multi-variant problems that balance various outcomes, a different type of digital assistant is needed. Lacking an industry term for this type of assistant, I have coined the term “Intelligent Digital Assistant” or IDA.

An IDA decision-making process is a dialog-driven, transparent, collaborative, and iterative process between machines and humans. The IDA not only presents recommendations, but it also prepares and presents the vital informational elements that informed its recommendation to the decision-maker. There is also shared accountability and learning for both the IDA and the decision-maker. Accountability is a fundamental concept for an IDA. Too often, we don’t review and understand the outcomes of our actions.

Finally, IDA technology is extensible. As more and more new data sources become available, we have opportunities to fine-tune our decision processes. However, too much information can result in “information overload” for decision-makers. The IDA can easily ingest the new information, incorporate it into model projections, and present a summary to the decision-maker. 

IDA technology, as a whole, is only beginning to emerge in the mainstream. However, the components needed to comprise an IDA solution are battle-tested. The dialog-driven interface has seen many successful implementations in the digital voice assistants discussed earlier. Graphical dashboards and data-driven alerts have been with us for quite some time and can be repurposed in the context of the IDA dialog to provide support for recommendations. AI and ML algorithms continue to improve with the advent of more/better data and more powerful processors.

The most well-known platform for developing IDA solutions is IBM Watson. Other major players in the space include Microsoft Oxford, SAS, and Google Deepmind. There are also a number of smaller players emerging, usually targeting specific use cases, but often times extensible to other problems. These include companies such as Quantexa, Quantellia, Lityx, and diwo.

No, good decisions don’t just happen. Organizations that adopt decision intelligence platforms will be in much better position than their competitors to navigate a very volatile and rapidly evolving business environment.

Related Articles

Back to top button