Score your role's resistance to automation — 8 factors
Eight evidence-based traits predict which roles stay valuable through 2028. Score yourself 0-2 on each, total out of 16, and read your result.
Major automation studies — Oxford's future-of-work research, McKinsey reports, the World Economic Forum's Future of Jobs — converge on a short list of traits that make work hard to automate: handling exceptions, carrying tacit knowledge, persuading people, and working in unstructured physical environments. This manual method scores your current role the way those studies do. For each factor below, score 0 (never in my work), 1 (sometimes, a few days a month) or 2 (most working days).
The eight factors and how to score them
- Unstructured problem solving — every task has a different answer, not a repeatable rule.
- Social and emotional intelligence — you negotiate, coach, de-escalate or read a room.
- Tacit knowledge — a new hire needs months, not days, to learn what you know.
- Physical dexterity in messy environments — hands-on work, tools, unpredictable conditions.
- Managing edge cases — you handle the 5% of requests that break the system.
- Cross-functional judgment — you weigh finance, legal and people trade-offs at once.
- External trust — customers or stakeholders choose to work with you personally.
- Continuous learning demand — your field changes meaningfully more than once a year.
The scoring table
| Points | What it looks like | Example |
|---|---|---|
| 0 | This factor rarely appears in my week | My job is 90% form filling and scheduled reporting |
| 1 | This appears occasionally, maybe monthly | I handle one tricky escalation per month |
| 2 | This is a core part of most days | I read and resolve ambiguous requests daily |
How to score honestly
Self-ratings almost always run optimistic by about one point. Before you finalize, ask a colleague who knows your work to read the eight factors and score you blind. Where your scores differ, trust the lower number — planning against the harsher estimate is what keeps the plan honest. A score you agree on with someone else is worth far more than one you graded alone.
Reading your total (0-16)
| Total | Result | What to do next |
|---|---|---|
| 0-5 | High automation exposure | Most of your tasks are routine. Over the next 6 months, deliberately move toward factors 1, 2 and 5 — ask for edge-case work and people work. |
| 6-10 | Moderately protected | Your role is partly safe. Identify which two factors you scored 0 on and engineer them into your work plan this quarter. |
| 11-16 | Strongly future-proof | Your value is judgment and relationships. Double down: document and teach what you know so it compounds into a moat. |
How it works
- Go through the eight factors and honestly score each one 0, 1 or 2 against a typical week.
- Add the numbers into a total out of 16 and find your band in the table.
- Pick your two lowest factors and write one concrete way to add them to your work this quarter.
They distill findings from major automation studies (Oxford's future-of-work research, McKinsey automation reports and the World Economic Forum Future of Jobs Report) into eight traits that consistently predict whether a task or role resists automation: judgment, empathy, tacit knowledge, physical dexterity, edge cases, cross-functional trade-offs, external trust, and learning demand.
No. It means the routine part of your role is exposed. The practical response is to deliberately move your work toward the factors you score zero on — for example, taking on edge-case handling or coaching teammates — rather than waiting for a layoff notice.
Yes, honestly. Physical dexterity only protects roles like maintenance, field service, healthcare hands-on work and skilled trades. An office worker who scores it 0 and scores high on judgment, edge cases and relationships is still strongly future-proof.
The replacement risk test scores your daily tasks (how routine they are). This page scores the enduring factors of your role itself. Use the risk test first if you want a task-level read; use this page if you want to decide which direction to grow.
Every 6-12 months, or whenever your role changes substantially. A useful score should climb over time as you deliberately add higher-resistance work; a flat score means your responsibilities are drifting toward routine.
Who this tool is for
Use this when the question is not will my tasks be automated but which direction should I grow. People who have already taken the replacement risk test often come here for the second half of the answer: the first page reads your week, this one reads the role. It suits individual contributors and first-line managers in office roles, and it is also a reasonable exercise for someone considering a move between two job families.
It is not a job-matching quiz and it says nothing about whether you are good at your job. Two people with the same title can score differently because they have shaped their work differently, and that difference is exactly what the page is designed to show.
How to read your result
Eight factors, each scored 0, 1 or 2 from the frequency in a typical week. The total runs from 0 to 16. The bands are fixed: 0 to 5 means high automation exposure, 6 to 10 moderately protected, 11 to 16 strongly future-proof.
The total alone will mislead you. Read the two or three zeros first. A score of nine built from zeros in physical dexterity and continuous learning describes a different situation from a nine built from zeros in external trust and cross-functional judgment. In the first case the missing pieces are hard to add quickly; in the second, a project or a rotation can move the number within a quarter.
Two honest notes. First, a zero on physical dexterity is normal for desk work and does not sink the total, because the remaining seven factors carry the score. Second, the factors describe the role as you currently perform it, not as your job description lists it. If you have quietly taken on escalation handling, score that factor for the work you actually do.
Finally, the page offers no prediction about 2028. The framing is a planning device: it tells you which kinds of work tend to hold value and lets you compare your week against them.
A worked example
An office manager at a 60-person manufacturing firm scores her role across a normal month. Unstructured problem solving gets 2: suppliers cancel, rooms double-book, and each week brings a new exception. Social and emotional intelligence gets 2: she calms internal disputes and reads the mood before a site visit. Tacit knowledge gets 1: a new assistant needs a few months to learn which shortcuts are acceptable. Physical dexterity gets 0: the job is desk-based. Managing edge cases gets 2: she handles the requests that fall between departments. Cross-functional judgment gets 1: she weighs facilities cost against staff complaints but rarely touches legal questions. External trust gets 1: two long-standing suppliers ask for her by name. Continuous learning gets 0: the software stack has not changed in two years.
Total: 9, moderately protected. The pattern is informative: the two zeros are learning demand and physical work, and the first is fixable while the second is not relevant to her career. Her plan for the quarter is to own the rollout notes for the new visitor system, volunteer for the cross-department scheduling dispute, and join the vendor renewal meetings so external trust moves from 1 to 2. If those three hold, a rescore in six months should land at 11 or 12 without changing employer.
Common mistakes
Scoring how well you perform a factor rather than how often it appears is the main one. The scale is about frequency, so a competent report writer can still score 0 on unstructured problems if the reports follow a template.
Office workers sometimes award themselves a point for physical dexterity because they stand at a standing desk, or because they carry boxes twice a year. That inflates the total without changing the underlying exposure. The reverse mistake is refusing any credit for judgment because anyone can be nice; the question is whether difficult people decisions land on your desk.
Comparing totals across different job families is not meaningful, and neither is rescoring every month. Once or twice a year, or after a real change in scope, is enough. And an eleven is not immunity; it is a signal about where to keep investing.