Over the years, I’ve noticed data scientists tend to fall into two camps: task-oriented and result-oriented. To make this concrete, let me introduce you to Jack and Jane.
Jack is task-oriented. He follows the model building process to the T without pausing and reflecting on the results. He creates beautiful plots, but doesn’t take time to think about what they reveal. Every stage of the pipeline is a box to check, and he checks them meticulously. Jack throws a punch in the air when the data shows a trend he suspected all along. He often complains about messy data. He’s efficient and thorough, but in his focus on the process he can lose sight of the bigger picture. He likes to work alone, and sees meetings with stakeholders as distractions — time taken away from “real” work.
Jane, on the other hand, is result-oriented. She treats process as a means, not an end. For her, the point of data science is to create something that matters — something that changes a decision, a strategy, a business outcome. She trusts data, but also likes to verify its veracity. She questions results that align too neatly with her priors, and digs deeper into ones that don’t. Jane collaborates. She listens to both technical and non-technical teammates. She values the messy business context and understands that data will always be messy. Jane knows the work isn’t successful until it’s adopted and acted upon.
Both types exist, and most of us probably have a bit of Jack and Jane in us. But if you want to create lasting value, my advice is: be more like Jane.
Because Jack risks being stuck in as a cog: valuable but replaceable. While Jane grows into someone who shapes decisions, not just someone who executes tasks.
Efficiency is good. But impact is better.