Being a Commercially-Minded Data Scientist
Getting you from good to great.
Background
In this article, I will refer to data analysts / data scientists as data practitioners. It is often said that what differentiates a good data practitioner from a great one is their commercial mindedness. It is no longer sufficient for data practitioners to simply code or model well, but to be able to find insights and create solutions that directly impact the business’ bottom lines. But what exactly does it mean for a data practitioner to be “commercially-minded”? And how would they go about acquiring this “skill”? Is it even a skill?
My 111th article unpacks what being commercially-minded means for data practitioners, and how they can hone this ability through the right kind of exposure. This topic is increasingly important in the Age of AI as data practitioners find their compute skills encroached upon.
(I write a weekly series of articles where I call out bad thinking and bad practices in data analytics / data science which you can find here.)
Commercially-Minded Isn’t A Skill
In general, being commercially-minded means having a strong awareness and understanding of how a business generates revenue, manages costs, and ultimately makes a profit. For a data practitioner, being commercially-minded extends to having an understanding of the key drivers of the business P&L, and the mechanisms of the organisation’s decision ecosystem that make things happen. They should also have sufficient appreciation of the competitive landscape in which the organisation operates.
Being commercially-minded isn’t a skill. It is a mindset, an attitude, a perspective. It’s whether the data practitioner considers commercial outcomes as inherently part of their job scope. They must wake up each day thinking about value extraction. There are obviously certain skills and competencies that support being commercially-minded. Of course, having business financial literacy is essential, but this knowledge is very easy to acquire, considering the intellectual and technical prowess of the data practitioner. But probably the most important competency underlying commercial-mindedness is for data practitioners to ascend from their stronghold of computational-thinking to complexity-thinking — the ability to see things as interconnected pieces, the ability to anticipate next steps and what-if situations, the ability to embrace uncertainty and even ambiguity. Complexity-thinking is essential to navigate the non-linear processes towards value extraction. Complexity-thinking is irreplaceable by AI (at present).
To Be Commercially-Minded
Being commercially-minded starts with accountability. The reality is that many data practitioners don’t believe that being commercially-minded is part of their job scope. For example, they might believe that their role is building the best predictive model in terms of reducing output uncertainties as much as possible, instead of whether the model output can be meaningfully acted upon to deliver incremental value. They might believe that they should be spending their time designing dashboards that visually amplify information signals, instead of how useful those information signals are in leading to corrective actions. They inherently believe that value extraction lies with those who are in receipt of the data practitioner’s outputs.
When I was given the opportunity to run the data practitioner department in Citibank Singapore more than 2 decades ago, having spent some time as a junior analyst in that team during my formative years, one of the first things I did was to assign P&L targets to my senior analysts’ performance KPI. That move was an intentional signal of things to come, and it made all the difference. There were some who didn’t like the new accountability and left. Those who remain chose to embrace it, and figured how to work with it. They became more proactive in their stakeholder engagement, more consultative, more directed in shaping P&L achievements.
And that’s exactly the DNA make-up of a commercially-minded data practitioner. They don’t ask “What models do you need?” but “How much incremental revenue do you need to generate?” and “What decisions and actions do you intend to take?”, then shortlist the range of solutions that could yield that, and present these options for further stakeholder discussions and inputs. They embrace the complexity of how solutions might evolve. They prioritise sufficiency over completeness / perfection. They look for connected friction points that might benefit from automation or uncertainty reduction. They look for re-usable assets to speed up the work. They ask the question “If this is successful, where do we go from here?”
Ultimately, being commercially-minded is a behaviour practice. It requires discipline, consistency, and commitment. Every stakeholder interaction, every business-as-usual activity, every problem-solving request, is an opportunity to practice … to explore …to reinforce.
Conclusion
If you are a data practitioner, you can’t “go to class” to learn how to be commercially-minded. You’ve just got to throw yourself over the deep end. Start by holding yourself accountable for a narrow set of business outcomes, even though it may not be in your performance KPI. Make it a point to deeply understand the business domain and decision ecosystem affiliated with that set of narrow business outcomes. Spend time thinking about how you can shift the needle on those outcomes. Test those ideas. Incremental effort every day. Incremental achievements every day.
