How I learned about Decision Intelligence and developed my own Decision Intelligence Navigator
Decision Intelligence: The Early Days
Decision Intelligence is a relatively new term and only became prominent in the last couple of years. However, one of the first ones to use the term was the famous management thought leader and professor James G. March in his 1994 book “A Primer on Decision Making: How Decisions Happen”. For whatever reason, the term Decision Intelligence didn’t become popular then and although the term popped up from time to time between then and now in a few publications and concepts, the term only became more widely known when Google started to use it and established the role of a Head of Decision Intelligence / Chief Decision Scientist in 2018 represented by Cassie Kozyrkov.
After March’s discussion of the term Decision Intelligence, there weren’t too many others using it for a long time. In 2002, Uwe Hanning wrote a book with the title “Knowledge Management + Business Intelligence = Decision Intelligence”. In 2004, Franklin Maxwell Harper used Decision Intelligence in the context of the possibilities that data warehousing was offering at that time and in 2007 two papers from XS Chang et al. as well as from Akihiro Inokuchi et al. mentioned the term Clinical Decision Intelligence. Interestingly, both author teams had very strong connections to IBM.
One of the strongest promoters of the term Decision Intelligence in the last couple of years (long before even Google started to apply the concept internally in 2018) is Dr Lorien Pratt, Chief Scientist and Co-founder of Quantellia LLC. She is doing a great job in making the concept of Decision Intelligence known to the wider public and I agree with her in many (not all) aspects of what Decision Intelligence is and can achieve.
Decision Intelligence: My Personal Path
I started working on the concept of Decision Intelligence around 2012 (first calling it “Industry Intelligence”) when I started to teach “Strategic Management” to MBA students at the University of St.Gallen in Switzerland. While trying to explain in 5 days (10 x 1/2 Day) of lectures how you create competitive advantages by applying different strategic analysis frameworks, drawing the right conclusions and acting upon it, it became increasingly evident to me that the factor that really mattered for successful companies was good decision-making. Obviously, this was not a breakthrough insight at all but it got me started to think how some kind of “Decision-based View” of strategic management could help my students to better understand how competitive advantages are created in reality. From such as Decision-based View of strategic management it was then only a small step to the concept of Decision Intelligence (at least for me).
Some first thoughts about Decision Intelligence that Adithya Vasudevan and I developed around 2013 are visualized in the next exhibit.
I then developed my first Decision Intelligence Navigator(TM) in 2013 (Version 1.0). It’s actually a simple collection of the four key elements that I had identified as essential parts of what I had then defined as Decision Intelligence (1.0).
Decision Intelligence (1.0) (in a management context) is the FIT between the intelligence (i.e. data, information, knowledge, insights) requirements of an executive and the total of an organization’s intelligence gathering & processing capacities.
The four elements of the Decision Intelligence Navigator 1.0 were labelled as “Intelligence Mindset”, “Framework Proficiency”, “Intelligence Access” and “Decision Proficiency”. I will describe each of the four elements and the latest versions of the Decision Intelligence Navigator at a later stage when comparing it with other approaches such as Google’s.
One of the first large companies that tried to leverage the concept of Decision Intelligence was Mastercard. My research has shown that the company had tried to protect the term through a trademark for several years but the company never got it granted. That’s why they call some of their services now MASTERCARD Decision Intelligence(TM). However, in the press release shown below, they actually use Decision Intelligence as a term with a trademark. A trademark that they never got — at least to the best of my knowledge.
When I realized in 2014 that you can’t protect the term Decision Intelligence (which makes perfect sense given that not even Mastercard was successful), I thought about how I could demonstrate my personal thought leadership in this area and as a consequence I:
a) renamed my “Industry Intelligence Ptv. Ltd.” 2014 into “DI-Decision Intelligence Ptv. Ltd.” in Switzerland and,
b) got the trademark for the Decision Intelligence Navigator.
I then also started to give public lectures and published the first papers on Decision Intelligence based on what I thought really matters.
I also would like to mention that I was following Dr Lorien Pratt from the very beginning and that her promotion and development of the Decision Intelligence concept was a very important influence. However, when I started to offer the first course on Decision Intelligence at the undergraduate level of the University of St.Gallen labelled as “Dealing with Uncertainty in Dynamic Markets” in 2014, I realized that students need some sort of clear visualization — a NAVIGATOR — to better connect the dots. Over a couple of years, I came up with the following Decision Intelligence Navigator which I use nowadays to teach students and executives alike the key ideas of what I believe Decision Intelligence is all about.
The Decision Intelligence Navigator consists of four major elements:
#1: Decision Context: Whether it’s business or any other context, a decision-maker needs to be clear about the specific DECISION CONTEXT that he/she faces (i.e. which decisions to make) and clearly define the QUESTIONS that she/he needs to get answered in order to make the decision(s).
#2: Framework Proficiency: Once you have identified the questions in #1 one needs to identify the FRAMEWORKS that are best suited to answer the questions. It could very well be that you need several frameworks to answer a single question and one framework can provide (partial) answers to several questions.
The more frameworks you know, the more framework-proficient are you!
#3: Intelligence Access: Once you have identified the most relevant frameworks in #2, you need to create ACCESS for you or others to the relevant INTELLIGENCE. The understanding of intelligence as a term is that it comprises data, information, knowledge or insights.
The FIT between the ANSWERS (#1) you need (to make a decision) and the most suitable FRAMEWORKS (#2) and required INTELLIGENCE ACCESS (#3) is what I then defined as DECISION INTELLIGENCE (3.0).
The result of DECISION INTELLIGENCE (i.e. the fit between the answers you need and the most suitable frameworks and best intelligence access) is INSIGHTS. Insights are defined as DECISION-RELEVANT KNOWLEDGE (i.e. what you need to know to make a decision — more on that in another article).
#4: Decision Proficiency: Once you have created the insights from the FIT between (#1 and #2 & #3) you need to make sure that the decision-relevant knowledge (i.e. insights) is translated into decisions avoiding being too much influenced by personal preferences, opinions or any sort of bias.
When I worked with executives on Decision Intelligence in the past, they all (finally) understood the logic of the concept and its importance to improving the competitiveness of their companies or their own careers but I often heard that they don’t know how to implement it or win over their colleagues in their companies.
This situation changed suddenly mid-2018 when Google under the leadership of Cassie Kozyrkov started to apply its own Decision Intelligence approach.
Decision Intelligence: A New Era with Google
When Cassie Kozyrkov became Google’s first Chief Decision Scientist (or Head of Decision Intelligence) in March 2018, the concept of Decision Intelligence suddenly became more widely accepted and all I had to do in my executive courses was to compare my Decision Intelligence Navigator with Google’s understanding of Decision Intelligence.
The key elements of these three definitions of Google’s Decision Intelligence understanding are the following:
- Decision Intelligence is an interdisciplinary challenge.
- It combines applied data sciences with social/managerial sciences but also integrates aspects of neurosciences & psychology.
- The final objective is to make better decisions and take actions.
I do not want to repeat the basic description of the Decision Intelligence Navigator from above but when I teach Decision Intelligence to students or executives we always work on the following challenges:
- Are you really clear about what the DECISION CONTEXT is?
There is no point in gathering or analyzing intelligence (data, information, knowledge) if you have not clearly defined who the decision-maker is, what kind of decision-making problem she/he is facing and what specific questions you (still) need to answer to make the decision.
- In order to achieve the best level of Decision Intelligence, you need to create the best possible FIT between the open QUESTIONS (#1) and the most suitable FRAMEWORKS (#2) and INTELLIGENCE ACCESS (#3).
That’s why we then work on the students’ and executives’ Framework Proficiency across disciplines (management, sociology, engineering, law, biology…) as well as the Intelligence Access they and their organizations have — This includes not only big data (analytics) in the form of data science / machine learning as Google emphasizes it but also the potential of small data (e.g. expert opinions) and other forms of data processing (e.g. ethnography) because not every question can be answered by big data and/or machine learning.
- Once you have created this FIT (i.e. Decision Intelligence) in the form of insights (i.e. decision-relevant knowledge) at the output, you (as decision-maker) still need to turn these insights into actual decisions and actions (the outcome). The most important challenge here is to avoid unnecessary biases and be aware of your own opinions (which are basically filters that let you differently weigh the insights you created).
I wrote an article on LinkedIn some time ago about what we need to consider when we look at the “value chain” from data to actions. Besides focusing on the differences between data, information, knowledge and insights (i.e. different forms of intelligence), it puts an emphasis on the role of different filters including concepts like wisdom, opinions and objectives.
After Google started to promote the Decision Intelligence concept it didn’t take long until other companies followed with slightly different terms but basically talking about the same. Some of the most prominent examples are Cognizant, IBM and Alibaba. While Cognizant talks about Intelligent Decisioning, IBM talks about Decision Intelligence in many presentations but calls its own product offering Decision Optimization. Alibaba was also quick with picking up the term but besides having a Decision Intelligence Lab they have gone their own way to promote their services in this area.
Meanwhile, there are literally hundreds of other companies — mostly tech companies that want to sell their AI/ML solutions to clients — talking about Decision Intelligence but almost none of them actually emphasizes the two first elements of the Decision Intelligence Navigator:
- DECISION CONTEXT (#1): Making sure you know which problems to solve and ask the right questions.
- FRAMEWORK PROFICIENCY (#2): Making sure that you apply the best possible frameworks/models and try to give answers to the questions (#1) from all relevant perspectives.
In order to emphasize this point in my lectures and executive coachings I often use the following quote from Albert Einstein:
As we still struggle to clearly define what Decision Intelligence actually is, what it includes and how it's different (or not) from Intelligent Decisioning or Decision Optimization, another major step for the broader acceptance of the concept was the inclusion of Decision Intelligence or Decision Intelligence Engineering into Gartner’s Top 10 Trends in Data and Analytics in 2020 and 2021.
In a nutshell, there currently exist dozens if not hundreds of different definitions or interpretations of the term Decision Intelligence and what it consists of — very similar to other popular management terms such as strategy or business models. We all know what we mean when using these terms but the details are normally only important when you really need a strategy or a business model. I believe we currently face the same situation with Decision Intelligence. We have all understood that in data-driven economies & knowledge societies, it is finally our decisions that truly make the difference. However, I believe that we currently put too much emphasis on the data science aspects of decision-making as compared to the framework/model and especially the question elements.
Let’s not forget that besides data science (big data, AI/ML), we have other tools at hand to make good decisions. If you want to achieve DECISION INTELLIGENCE you need to master all tools in the box.