Data science thrives when business insight meets technical expertise. That balance is harder to achieve than it should be, and the experience of this data scientist, three years in the field, two roles at one large company, proves the point. They moved from a consulting-heavy position to a more technical one, worried about losing their engineering edge. Now they find themselves cut off from actual business stakeholders, with a manager who reserves those connections for themselves and a single lead. This is not a personal failure. It is a structural flaw in how many organizations design data science teams.
The practical consequence is clear: technical skill without business context is a machine with no purpose. You can build a flawless model, but if you do not understand what the business needs, what questions it is asking, or what constraints it operates under, that model sits on a shelf. Conversely, pure consulting-focus without production engineering leaves you unable to deliver at scale. The false choice between these two paths is what traps so many practitioners. The problem is not that data scientists lack ambition. The problem is that companies treat business alignment and technical depth as separate career tracks, when they are actually two legs of the same journey.
What this means for you as a data scientist is that you must actively resist being siloed. If your manager does not support cross-functional exposure, that is a signal worth heeding. Look for teams where the culture rewards asking questions of both the data and the people who own the decisions. Seek out roles where the job description explicitly mentions working with stakeholders, not just building pipelines. The best data science work happens at the intersection, where a deep understanding of regression or clustering meets a real grasp of revenue cycles, customer behavior, or operational risk. If your current environment makes that intersection impossible to reach, then yes, start looking. Not because you are failing, but because the structure around you is.
The takeaway is not that you must choose between technical depth and business fluency. It is that you should not have to. Organizations that force that choice are wasting their talent. And data scientists who accept it are limiting their own impact. So push back where you can, and leave where you cannot. The goal is not to be the best coder in the room or the best translator of business needs. The goal is to be both, because that is where value actually gets built.