Verified against ChatGPT · 2026-08-12
Build a career roadmap as a decision tree, not a straight-line plan
Produces a multi-year career plan structured around the actual decision points and forks ahead of you — what has to be true to unlock each path — instead of a single confident linear timeline that ignores how uncertain the next move actually is.
The prompt
Ready to copy — highlighted parts are example details you can swap.
You are helping me build a career roadmap for the next 3-5 years — but I don't want a single confident straight-line plan, because I genuinely don't know which of two or three directions is right yet, and a fake-precise timeline would be more misleading than useful. CURRENT POSITION Mid-level data analyst at a retail company, mostly building dashboards and ad hoc reporting. DIRECTIONS I'M CONSIDERING 1) Move toward data science / ML, 2) move into analytics management, 3) move into a product-adjacent analytics role at a startup. WHAT WOULD MAKE ME CHOOSE ONE OVER ANOTHER Whether I enjoy people-management (untested), and whether I can get hands-on ML project experience in my current role within a year. CONSTRAINTS Can't take a pay cut for at least 18 months; based in a city with few ML-heavy employers. Build this as a decision tree, not a single path. For each candidate direction, identify the near-term move that's common to it and at least one other direction — the thing worth doing regardless of which path turns out right — versus the moves that are direction-specific and would need to be reversed or abandoned if I picked differently. For each direction, name the specific signal or decision point, roughly timed, at which I'd have enough information to commit or rule it out — not a vague "see how it goes" but a concrete thing to observe (a project outcome, a skill gap that either closes or doesn't, feedback from a specific type of stakeholder). Where two directions genuinely trade off against each other given my stated constraints, say so plainly rather than implying I can pursue both fully at once. If one of the candidate directions is clearly weaker given my stated deciding factors, say that directly rather than treating all options as equally viable out of politeness. OUTPUT FORMAT 1. A short list of "no-regret moves" — things worth doing under any of the directions, with rough timing. 2. One subsection per candidate direction: the direction-specific moves, the decision point that would confirm or rule it out, and roughly when that signal would appear. 3. A one-paragraph honest read on whether any direction looks weaker than the others given what I've told you, or whether it's genuinely too early to tell.
Customize
Optional — swap in your own details for the highlighted parts above.
Why this works
Explicitly asking for a decision tree instead of a linear plan preempts a specific default behavior: asked for a "5-year career roadmap" without this framing, GPT-5.1 tends to produce a single confident year-by-year narrative that reads well but papers over the real uncertainty of an unresolved fork, which is actively counterproductive when the honest state of things is that the user doesn't yet know which path is right and forcing false precision would just produce a plan that gets abandoned at year one. Separating no-regret moves from direction-specific moves gives the plan something actionable to do this week regardless of which fork gets taken, which matters because career plans that are entirely conditional on a future decision tend to produce paralysis in the present — there's always a reason to wait for more information before acting. Requiring a concrete, observable signal for each decision point rather than a vague "see how it goes" addresses the fact that unstructured self-reflection about career fit is notoriously unreliable; a person rarely discovers whether they like people-management by introspecting about it, but they do find out by managing one project and noticing whether it energized or drained them, so the prompt forces the model to name that kind of concrete test rather than deferring the whole decision to an unspecified future feeling. The instruction to say plainly when one direction looks weaker matters because an AI model's default politeness bias tends toward presenting all user-suggested options as equally viable, which is a disservice when the user's own stated constraints — a fixed geography with few ML employers, an income floor — already rule one path down; naming that directly is more useful than diplomatic neutrality.
What you get back
No-regret moves: pick up one hands-on statistical modeling project in your current role within 3 months regardless of direction — this serves both the data-science and analytics-management paths. Direction: Data science / ML — Decision point: within 6-9 months, whether you can land a genuine modeling project (not just dashboards) internally. If your company has no such project pipeline, this path likely requires a lateral move to test, which conflicts with your no-pay-cut constraint for the next 18 months...
Verified against
ChatGPT GPT-5.1 · 2026-08-12
Changelog
- 2026-08-12 — Initial publish, verified against ChatGPT GPT-5.1.
Need this built into your business?
If a prompt isn't enough — what Scult builds, built and maintained for you — that's Scult's day job.
EXPLORE WHAT SCULT BUILDS
