Many organizations today are investing heavily in data projects to improve data-driven decision making. But paradoxically most of these projects don’t consider decisions explicitly. Data projects usually begin with data capture before understanding what data will be useful, for what action and for what benefit. Without understanding the decision making that will improve business performance and deliver the desired impact, data investments fall short of their promise.
Decision modeling is an industry standard* requirements technique that models decision making. By starting with the decision, data projects can determine what data they need to collect, what data must be sourced from outside the organization, at what granularity and other important details. As decisions are modeled, goals, processes, and data come together, and it is clear where to include knowledge and expertise, and where there are opportunities for a BRMS and advanced analytics.
What Decision Will This Help You With?
A decision model is built iteratively, by asking key questions:
- What decision will this help you with?
- How will you tell you are making better decisions with this data?
- How do you make this decision today and how much are you willing to change?
- What would you want to know to make this decision more accurately?
Decisions models are easy to understand graphical representations showing how a decision is made, along with the information and knowledge required to make an effective decision. Information Sources are generally data sources and Knowledge Sources could be well-defined Analytical Models, Business Rules and Algorithms or placeholders for human expertise and similar informal knowledge sources. Each element on the Decision Requirements Model can be further described with other relevant characteristics that describe the context.
Agile, Clear and Actionable
Decision requirements models can easily be expanded in a series of iterations, allowing progress to be made in an agile approach. Projects adopting decision modeling have shorter implementation times than those with a more general data approach, saving time and resources. Furthermore, decision modeling helps avoid rework that may be necessary when business drivers are incorporated after the fact.
The goal is to articulate all components of decision making without getting distracted with technology or implementation details. These high level requirements can then be used to define data requirements, operational dashboards and reporting requirements, as well as scope out technology choices, implementation phases and operating specifications related to governance and compliance. Focusing data projects on improving decisions makes the connection from data to data-driven decision making clear and actionable.
*Decision modeling is described using the Decision Model Notation (DMN) open standard from the Object Management Group (OMG), an open standards consortium.