Why Product Sense is one vastly underrated skill, even when compared to machine learning, statistics, or coding.

Martin Leitner
2 min readDec 22, 2022
Photo by Shahadat Rahman on Unsplash

Product Sense is a Superpower.

A common question I get asked is: “What skill is a significant differentiator for a Data Scientist?” Here are some thoughts in response to this question.

In my experience, a strong product sense is an essential factor that makes a data scientist stand out. This generally shows up in the following:

  • Actionable outcomes
  • Scalability
  • Efficiency and speed

My opinion on the importance of product sense sometimes surprises those who ask me, expecting me to call out elements within machine learning, statistics, or coding. And don’t get me wrong, those are undoubtedly important, but product sense is one vastly underrated skill.

Let me explain why I believe this to be significant.

1.) Actionable outcomes: This may seem obvious to most people — you can design a highly complex model, but if you can’t put it into practice, you won’t see any impact or tangible results. Understanding the stakeholders, their pain points, working practices and user flow, and thoroughly understanding any limits, whether hardware, software, or human resources, are all crucial considerations before entering the modeling phase. You will be able to contribute much more to the scope of the work and boost your status as a genuine thought partner to the stakeholder. From my observations, the “if you build it, they will come” adage seldom pans out.

2.) Scalability: How might the work help other workstreams (current or future), and if so, what does it take to set a foundation that allows for this broader application down the road?

3.) Efficiency and speed: Data scientists must make several choices on how to proceed in their work, requiring them to be decisive as they learn and adapt. The earlier you grasp the challenge and how a solution may emerge, the more methodically you will be able to progress toward the intended objective, minimizing scope creep and enhancing your feature engineering due to a more profound intuition of linkages back to the business.

As always, curious about your experiences and thoughts. Please let me know.

Want to connect?

--

--

Martin Leitner

Head of Data Science @Mars | creating game-changing impact through customer-centric, data-first strategies | triathlete, creative & disruptive thinker