4.1 The Temporal Shift (Time Horizons)
The single hardest adjustment for a newly promoted Staff DS is time horizon.
As a Senior DS, you were an “Agile Native.” You lived in 2-week sprints. Your dopamine loop was closed every Friday when you merged a PR. Your visibility was the quarter.
As a Staff DS, you must divorce yourself from the sprint.
Your job is no longer to “close tickets.” Your job is to own a thesis about where the product needs to be in 18–24 months, and shape the data strategy backward from that future. If you are only solving problems that can be finished in two weeks, you are not doing Staff work; you are doing maintenance.
Most teams plan forward: “What can we build next sprint?”
Staff DS plan backward: “If we want to be an AI-first decision engine in 2028, what infrastructure must we build today?”
You are the custodian of the long arc.
- The Shift: While the team argues about optimizing a query for next week’s report, you are arguing about the schema change that will prevent a data swamp next year.
- The Unwritten Rule: You will often look “unproductive” in the short term. You are advocating for “invisible work” (refactoring, metric governance, platform migration) that has zero immediate ROI but prevents the org from hitting a wall in 12 months. You must have the political capital to survive the “valley of despair” before the payoff.
Senior DS optimize for speed: “How quickly can we get significance?”
Staff DS optimize for structural truth, even if it is slow.
You design the bets that take 6–12 months to pay off.
- Example: A Senior DS runs a button-color test (1 week). A Staff DS runs a long-term holdout group (6 months) to measure the true incrementality of the loyalty program.
- The Trade-off: You have to look a VP in the eye and say, “We will not have an answer for this quarter. But in Q4, we will have the right answer, and it will change our strategy for the next three years.”
In a 2-week sprint, being wrong is a bug. You fix it in the next sprint.
In a multi-year arc, being “wrong” is the default state for the first 6 months.
You are betting on a future that has not happened yet (e.g., “We need to move off cookie-based attribution before Chrome deprecates them”). For the first few months, you will sound like an alarmist. The data will not support you yet. The metrics will not move yet.
- The Skill: You must be comfortable being directionally correct while being precisely wrong.
- The Unwritten Rule: The Senior DS fears being wrong about a number. The Staff DS fears being right about the wrong thing. You accept the “ambiguity tax” today to buy clarity for the future.
Use this heuristic to audit your calendar:
Horizon | The Role | The Question |
2 Weeks | Senior DS | “How do I ship this analysis by Friday?” |
3 Months | Lead DS | “How do we hit the Q1 OKR?” |
18 Months | Staff DS | “If we hit the Q1 OKR but ignore this trend, are we dead in 2 years?” |
4.2 Power, Politics, & The Shadow Org Chart
If you flinch at the word “politics,” you are not ready for Staff.
Junior engineers think politics is about sucking up or manipulating people. Staff engineers know that politics is simply the mechanism for resource allocation in an environment of scarcity.
You do not have the authority to order people to do things. You have zero headcount. You cannot fire anyone. Yet, you are expected to steer a ship of 50+ people. To do that, you must master the “illegible” layers of the organization.
On Workday, you report to a Data Science Manager or Director.
In reality, you report to the problem.
At L6, the principal–agent relationship shifts. Your manager is no longer your “boss” who hands you tasks; they are your “support node” who unblocks you.
- The Shift: A Senior DS asks their manager, “What should I work on?” A Staff DS tells their manager, “Here is what the business needs, and here is how I am going to solve it. I need you to cover me.”
- The Unwritten Rule: You must be willing to disappoint your manager in the short term to save the business in the long term. If you prioritize “making your manager happy” over “fixing the structural error in the roadmap,” you are operating as a Senior, not a Staff.
The best manager I had (among several) was a Director at Meta who said:
“You’re the expert, you tell me what the biggest problems in this domain are and what your solution is. And how I can help you in terms of elevating the problem, aligning people outside of your org, and resourcing bottlenecks from other teams.”
They also said:
“Don’t give me status updates. Tell me your blockers and how I can help.”