Generative Artificial Intelligence (GenAI) is revolutionizing our approach to data analytics and we are just witnessing the beginning of its impact. How, then, will GenAI influence both the demand for Decision Scientists and the evolution of this role within organizations?

GenAI has notably enhanced the efficiency of data scientists by speeding up code generation and model build processes. Complex data transformations, once extremely time-consuming (it has been estimated that data scientists spent 70-80% of their time on data wrangling and transformations), now take a fraction of the time.  Additionally, GenAI levels the playing field for practitioners, enabling even average programmers to achieve top-tier results in programming languages like Python, R, SAS, and SQL. This significant time saving importantly shifts the focus of the data scientist to other critical tasks, namely exploratory analysis, model building and optimization, and interpretation of results.

With this re-allocation of time, the data scientist now has additional capacity to develop more in-depth, innovative — and sometimes inherently more complex — data science solutions. With GenAI acting as a heuristic guide during the exploratory and model build phases, data scientists can cover more analytical ground more quickly and dive deeper into the important work of enhancing decision-support. Data scientists can now test a wider array of statistical methods and more keenly direct their focus to the important work of creating expected value frameworks, for example, to extend model outputs to better anchor decision-support.  Such frameworks are critical to estimating the impact of different decisions and spend allocations (e.g. incrementality and lift) on marketing performance, even generating “what-if” scenario analyses and probabilistic simulations.

With these changes in mind, I believe that GenAI is poised to unveil a new echelon of skilled data science professionals in business, whom I refer to as Decision Scientists. This emerging role will bridge the strong technical skillsets of today’s data scientists with deeper business acumen, robust analytical thinking, and problem-solving creativity necessarily required to support data-driven decision-making.¹ Rate-limiting factors on performance once imposed on bright minds by technical constraints like mastering coding languages, for example, are lessening.  The time saved can now be used to better focus on insight generation (maybe now only 30% of time need be allocated to data wrangling?).  In this context, GenAI is a new, powerful tool that is helping Decision Scientists more easily shift focus from the science of inputs (data and analysis) to the science of outputs (recommendations and decisions).²

Much of the analysis that takes place in organizations today — maybe even the majority of analysis — never leads to clear, prescriptive recommendations to support decision-making.  This is not to say that these analyses are not important in certain contexts.  However, they are mostly summarizations of interesting facts and stats that seldom are mobilized into a strategic plan of action. This has been the “data paradox” of my entire career: while companies have access to more data than ever before, it’s rarely used to its full potential in strategic decision-making.  A wealth of reports and historical data too often leads to paralysis rather than action. Companies must move beyond mere insights to clear strategic actions, where decision-making is a scientific process, grounded in data, not driven by intuition.

In light of these challenges, the Decision Scientist role is to address the disconnect between data availability and its strategic use.  They are solely focused on using the full suite of data mining, statistical, or machine learning methods available to them to discover, quantify, and produce clear, strategic recommendations — even when the evidence is fuzzy — and close the “insight-to-action” gap.  They are not preoccupied with simply answering questions with data.  In the case of Marketing decision-making, in particular, they are obsessively focused on understanding incrementally and lift, which should be the central inflection point of nearly all decisions-making related to performance optimization.

In many ways, GenAI empowers Decision Scientists to forge deeper connections between ideas and data-driven solutions more easily and accurately, especially as GenAI capabilities mature in the coming years. While remaining a highly technical field — requiring proper education and training in data science — GenAI is transforming what we know today as data science into a much more exploratory, and even creative, endeavour allocating much more time in each day to doing the good work of directly supporting strategic decision-making.

In the coming years, stakeholder expectations for forward-looking predictive and prescriptive insights as a standard ad-hoc deliverable will increase dramatically.  The expectation will now be for the delivery of very precise recommendations based on different scenarios and associated levels of risk, in a much timelier manner than ever possible previously. 

1. Data Science vs. Decision Science: What’s the Difference? ↩︎
2. Since Data Scientists Don’t Do Business Decisions — Businesses Need Decision Scientists ↩︎