Decision Science — Between Cognitive & Data Science
A Very Practical Space In The Universe
Data science has all the buzz. Cognitive science less so, but it holds the airs of academia. Somewhere between lies the realm of the decision scientist, a very practical role that has never received much buzz or acclaim. It is a position that is all business.
Data scientists work with big data. It is not their exclusive landscape but you wouldn’t know that by anything in the popular media. They are the creators of machine learning, artificial intelligence, and infamous ‘algorithm’. It is a realm of PhDs, or at least it has been.
Cognitive scientists work with the human mind. Some trying to develop neural networks and computing that can rival it, others cranking an endless array of academic papers attempting to explain it. It is a multi-disciplinary field, but still one ruled by PhDs.
While a PhD in Decision Science is not unheard of, it is certainly not nearly as common a trait for practitioners as occur in these other areas. Many large corporations have created decision science teams. These teams grew out of decision support and planning & analysis teams (FP&A or BP&A) of the 90s. They are often tasked with the development of Decision Support Systems (DSS). Of the three functions, it holds the least acclaim and the most practical purpose.
Decision Support Systems
Perhaps one of the reasons Decision Science remains less visible than it’s cousin disciplines, the vagueness of its most common product. What is a DSS? It is a great question that lacks a truly great answer.
A DSS can take the form of a statistical model. Wait, isn’t that data science? It can be a simple mental framework. Wait, cognitive science? It can be as simple as a report or dashboard. Come on, that is business intelligence!? Or is sometimes labeled as an expert system. Software development? I said there was no great answer.
Like much of the data science & analytics landscape, decision science has been conflated and confused. I suppose there is nothing unique here. It is just a bit disappointing to see things labeled science and discipline be so poorly defined and subjective… but I digress.
Decision science, when done well and well defined, is a very practical field. It produces frameworks and models, systems and reports, policy and planning. This is not to say that cognitive science is never practical, though that seems like less of a defining feature. It is also not to say data science isn’t practical, though I would lean heavily on Industry 2.0 to carry that argument. It is to say that decision science is defined by its practicality. Or but differently, the AI in Decision Science is Actionable Intelligence.
Is Decision Science more multi-disciplinary? Perhaps. It all depends on how you define it. Do some companies lump the role of decision science into other areas? Of course — data science for one, but also strategy, finance, and even operational functions. Very few actually have a cognitive science function, but that may eventually change, as well.
Decision Science biggest differentiation from data science and cognitive science, beyond practical/actionable output, is its ties to the P&L. Cognitive science often captures aspects of game theory and economics, but being distant from business in general — it rarely captures the nuances of accounting, budget, ROI, and pro-formas. It is also closer to management than psychology.
Finally, Decision Science does have its own buzzword. Data-driven decision-making or DDDM has had its day in the sun. While it can certainly be connected to data science and cognitive science, it lives far more soundly with Decision Science.
DDDM ties decision science to the culture of a company. This is an element that, so far, neither data nor cognitive science have tried for. Most companies with a named data science team (and the rare few with a cognitive science group) view them as outside the culture of the company. To borrow an odd term that should have died off long ago — propellerheads.
Decision science teams may seem a bit nerdy to the rest of the business, but the level of distance that defines propellerheads is typically avoided. It is reflected in the language that these teams use — or don’t, as they tend to act as translators between departments. It is also somewhat necessary as distance works against outcomes and practical utility.
Whether your company has a decision science team, a data science team, or some other hybrid the decision science function is necessary — assuming anything practical is going to result from leveraging data. Sadly, that assumption is not universal. Many a company ignores this function and consequently wonders why they “waste” so much money on meaningless data manipulation.
Of course, now, you aren’t likely to make that same mistake! Thanks for reading!
