An Introduction to the Emerging Field of Decision Intelligence

What is Decision Intelligence?

Decision intelligence (DI) is a new discipline that combines data science, social science, and managerial science to help organizations make better decisions. As the world becomes increasingly complex and data-driven, traditional approaches to decision making are no longer sufficient. Decision intelligence provides a framework and set of tools for navigating this complexity and leveraging data, AI, and human judgment to drive optimal outcomes.

At its core, decision intelligence is about understanding and modeling the chain of cause-and-effect from decisions to outcomes. It involves mapping out the key decisions an organization needs to make, identifying the factors and uncertainties that impact those decisions, predicting the likely outcomes of different decision scenarios, and then optimizing the decision making process based on goals and constraints. By connecting decisions to outcomes in a rigorous, quantitative way, DI helps decision makers move beyond gut feel and guesswork to a more scientific approach.

Why Decision Intelligence Matters

Businesses and organizations today are grappling with unprecedented levels of complexity, uncertainty, and change. From shifting consumer behaviors to disruptive technologies to economic volatility, the pace and magnitude of change has never been greater. At the same time, the proliferation of data and advanced analytics tools means that organizations have access to more information than ever before to guide their decisions.

However, having lots of data is not the same as harnessing that data to make good decisions. In fact, research shows that more information and analysis can actually lead to decision paralysis and poorer decisions if not used effectively. This is where decision intelligence comes in – by providing a structured way to integrate both quantitative data and qualitative factors, collaborate across silos and stakeholders, and continuously learn and adapt based on outcomes.

The need for more intelligent, data-driven decision making is evident across virtually every domain. For example:

• In business, decision intelligence can help optimize marketing spend, pricing, product development, supply chain operations, and more to maximize profitability and competitiveness.
• In healthcare, DI can help inform clinical decision making, resource allocation, population health management, and policies to improve patient outcomes while controlling costs.
• In government, decision intelligence can help shape smarter policies and investments in areas like infrastructure, education, social services, environmental protection, and national security.
• In non-profits and NGOs, DI can help optimize limited resources and interventions to maximize impact on crucial global challenges like poverty, hunger, disaster response.

The common thread is that decision intelligence provides a way to tackle complex, high-stakes decisions in a more rigorous, scientific manner. By leveraging data and analytics alongside human expertise and values, DI aims to both improve individual decisions and make organizations smarter and more agile overall.

Key Concepts in Decision Intelligence

To understand how decision intelligence works in practice, it‘s helpful to break it down into some of its key components and concepts:

Decision Modeling: This involves mapping out the key decisions an organization faces, the factors that influence those decisions (data, assumptions, constraints), and the relationships between decisions. Decision models provide a visual representation of the decision process and make it easier to identify gaps, inconsistencies, and opportunities for improvement.

Causal Reasoning: Understanding cause-and-effect is central to decision intelligence. This means digging beneath surface-level correlations to uncover the mechanisms of how different factors actually influence outcomes. Causal models, often represented as causal diagrams or Bayesian networks, encode an organization‘s best understanding of causality to guide decision making.

Simulation and What-If Analysis: With a causal decision model in place, decision intelligence platforms can then simulate the effects of different decision scenarios and compare their likely outcomes. Through what-if analysis, decision makers can ask counterfactual questions, stress-test strategies against uncertainty, and find the optimal decision path to achieve their objectives.

Decision Automation and Augmentation: For decisions that are high-volume, time-sensitive, or have clear best actions based on data, decision intelligence enables varying degrees of automation, from simple decision rules to sophisticated AI agents. Conversely, for more complex, qualitative decisions, DI focuses on augmenting and empowering human decision makers by providing them with actionable insights and recommendations.

Continuous Improvement: Finally, a key tenet of decision intelligence is that decision making is not a one-time event, but rather an ongoing, iterative process. By closing the loop between decisions and outcomes, learning from experience, and adapting the decision model over time, organizations can continuously optimize their decisions and build a sustainable competitive advantage.

Intersection with Related Disciplines

Decision intelligence is a highly interdisciplinary field that builds upon and integrates concepts and techniques from various related disciplines:

Data Science and Analytics: Decision intelligence leverages the tools of data science, including data mining, statistical modeling, machine learning, and data visualization, to extract insights and make predictions based on data. However, DI goes beyond pure data analysis to focus on driving decisions and actions.

Artificial Intelligence and Machine Learning: DI uses AI and ML techniques to process large amounts of data, identify patterns and relationships, and generate decision recommendations. However, rather than fully automating decisions, DI often uses AI to augment human decision making and focuses on human-machine collaboration.

Behavioral Economics and Psychology: Decision intelligence incorporates principles from behavioral economics and psychology to understand and account for cognitive biases, social dynamics, and other "irrational" factors that influence human decision making. DI seeks to design choice environments and decision support systems that nudge people towards better decisions.

Management Science and Operations Research: DI applies many of the optimization, simulation, and decision modeling techniques from fields like management science and operations research. However, it also emphasizes the organizational and human aspects of decision making, seeing it as a social process as much as an analytical one.

Design Thinking and Agile: Decision intelligence shares with design thinking and agile methodologies an emphasis on fast iteration, user-centricity, and adaptation based on feedback. DI favors experiments, prototypes, and minimum viable products (MVPs) to quickly test and refine decision approaches in the real world.

Applications and Case Studies

Decision intelligence is already being applied across a wide range of industries and domains to improve decision making and drive better outcomes. Here are just a few examples:

• Retail: H&M, the global apparel retailer, uses decision intelligence to optimize its merchandising decisions, from what styles to stock in each store to how to allocate inventory across its network. By leveraging granular sales data, customer demographics, and demand forecasts, H&M can tailor its assortment to each local market while still maintaining a cohesive brand identity.

• Financial Services: Stripe, the payments processing platform, leverages decision intelligence to prevent fraud and manage risk. Stripe‘s machine learning models analyze hundreds of variables with each transaction to identify suspicious patterns in real-time, allowing it to block fraudulent charges with high accuracy. At the same time, human reviewers are kept in the loop to interpret edge cases and provide feedback to continuously tune the models.

• Healthcare: The Carilion Clinic, a major healthcare system in the US, has adopted a decision intelligence approach to optimize nurse staffing and scheduling. By analyzing data on patient census, acuity, staff skills, and more, the system generates optimal staffing recommendations to balance quality of care, nurse satisfaction, and cost efficiency. The platform also allows nurse managers to do what-if analyses to plan for different scenarios.

• Government: The city of Los Angeles has embraced decision intelligence to tackle homelessness, one of its most pressing challenges. The city‘s data science team has built predictive models to identify people at highest risk of homelessness based on service utilization data, allowing interventions to be targeted proactively. The team has also developed a geospatial optimization tool to inform the placement of new housing resources to maximize impact.

Trends and Future Directions

As an emerging field, decision intelligence is evolving rapidly, with new technologies, methodologies, and applications constantly pushing the boundaries of what‘s possible. Here are some of the key trends and future directions that are shaping the field:

Explainable AI: As machine learning models become more complex and opaque, there is a growing need for explainable AI techniques that can provide clear rationales for AI-generated decisions. This is particularly crucial for high-stakes decisions in areas like healthcare and criminal justice, where human oversight and accountability are essential.

Causal Inference: Causal modeling is a core part of decision intelligence, but traditional methods rely heavily on domain expertise and assumptions. Newer techniques from causal inference, such as causal discovery algorithms and counterfactual reasoning, can help automate the process of uncovering causal relationships from data, leading to more robust and adaptive decision models.

Real-Time Decisions: With the proliferation of streaming data and edge computing, there is an increasing demand for real-time, event-driven decision making. This requires decision intelligence platforms to be able to ingest and analyze data in real-time, detect significant patterns and anomalies, and trigger appropriate actions based on pre-defined rules or policies.

Collaborative Decision Making: Decision intelligence recognizes that decision making in organizations is often a team sport, involving multiple stakeholders with different expertise, incentives, and viewpoints. Emerging collaborative DI tools focus on facilitating this group decision making process through features like shared workspaces, discussion forums, voting and consensus building tools, and more.

Ethical AI: As AI becomes more deeply embedded in decision making systems, it is crucial to ensure that these systems are transparent, unbiased, and aligned with human values. Decision intelligence will increasingly incorporate ethical principles and constraints into its models and algorithms to promote fair, accountable, and socially beneficial decisions.

Getting Started with Decision Intelligence

If you‘re interested in learning more about decision intelligence or applying it within your own organization, here are some recommended resources to get started:

• Books: "Link: How Decision Intelligence Connects Data, Actions, and Outcomes for a Better World" by Lorien Pratt, "Thinking in Bets: Making Smarter Decisions When You Don‘t Have All the Facts" by Annie Duke, "Decision Intelligence For Dummies" by Pamela Baker.

• Online Courses: "Decision Intelligence Foundations" by Quantellia, "Introduction to Decision Intelligence" by Busigence Academy, "Decision Making in a Complex and Uncertain World" by MIT.

• Conferences: Decision Intelligence Summit, Predictive Analytics World, Data for AI Week, Chief Data & Analytics Officers.

• Blogs and Thought Leaders: Cassie Kozyrkov (Google), Eric Siegel (Predictive Analytics World), Lorien Pratt (Quantellia), Jian Gao (H&M).

Ultimately, the most important step is to start asking the right questions within your organization: What are our most critical decisions? How do we currently make those decisions, and how well are they working? What data and models could we leverage to make better decisions? What would a more intelligent, agile decision making process look like?

By framing decision making as a strategic capability and investing in decision intelligence tools and approaches, organizations can position themselves to thrive in an increasingly complex and data-rich world. The future belongs to those who can harness the power of data, technology, and human judgment to make smart decisions – and decision intelligence lights the path.