10 min read
How Leveling Works
For: Managers preparing a level request or promotion case
The six dimensions that decide whether a role is junior, mid, senior, principal, director, or executive.
Step 1 of 5. The six dimensions, in one minute.
Step 1 of 5
The six dimensions, in one minute
Six dimensions cover most of what makes one job bigger than another. Score the role on each, weight them, get a directional level.
Most leveling models size jobs along the same handful of dimensions. The labels vary a little. The shape doesn't. I picked six because the ones below cover most of the practical ground without becoming a check-the-box exercise.
| Dimension | Question it answers | Confused with |
|---|---|---|
| Scope | What does the role reliably own? | Workload, hours, tenure |
| Complexity | How ambiguous are the problems? | Volume of tasks |
| Autonomy | How independently does it operate? | Lack of supervision |
| Influence | Whose decisions does it change? | Likeability, network size |
| Knowledge | What depth of expertise is required? | Years of experience |
| Business impact | How material is the impact? | Visibility, prestige |
The six core dimensions
- Dimension
- Scope
- Question it answers
- What does the role reliably own?
- Confused with
- Workload, hours, tenure
- Dimension
- Complexity
- Question it answers
- How ambiguous are the problems?
- Confused with
- Volume of tasks
- Dimension
- Autonomy
- Question it answers
- How independently does it operate?
- Confused with
- Lack of supervision
- Dimension
- Influence
- Question it answers
- Whose decisions does it change?
- Confused with
- Likeability, network size
- Dimension
- Knowledge
- Question it answers
- What depth of expertise is required?
- Confused with
- Years of experience
- Dimension
- Business impact
- Question it answers
- How material is the impact?
- Confused with
- Visibility, prestige
Try one dimension
Scoring one dimension makes the rest concrete. The hard part is calling the close ones with evidence.
The dimensions sound abstract until you score one. Pick a real role you know well, then click through Scope.
Two roles, scored side by side
Side-by-side comparisons surface the dimensions doing the real work. The same total can come from very different combinations.
Two roles can have the same title and end up at very different levels. The dimensions show why.
- PM A: Scope 4, Complexity 4, Autonomy 4, Influence 3, Knowledge 4, Impact 4.
- Core average ≈ 3.83. Maps to P4 Advanced. Confidence: High.
- PM B: Scope 5, Complexity 5, Autonomy 5, Influence 5, Knowledge 5, Impact 5.
- Core average = 5.00. Maps to P5 Expert. Boundary fires at P5/P6 (the score is within 0.10 of the 4.9 threshold). Confidence: High.
- Same title. Different levels. The label was carrying weight the scope didn't earn.
Math format and band ranges are documented in `leveling-engine-reference.md`.
| Pattern | Shape | Result |
|---|---|---|
| Narrow scope, deep problems | Owns a small area but the problems inside it are genuinely hard. | P5 Expert |
| Balanced senior | All six dimensions sit in the same neighborhood. The pattern that maps cleanest to a level. | P6 Principal |
| Wide reach, simpler work | Owns a big surface area, but the work inside it is mostly known plays. | P6 Principal |
Manager and executive add a seventh dimension
Managers add People Leadership. Executives add Strategy Ownership. Same six core dimensions, plus the one their track requires.
Manager- roles add people leadership at 20% weight. Executive- roles add strategy ownership at 30% weight. The math is in the engine reference. The takeaway: if the role doesn't have meaningful people leadership or strategy ownership, picking that is going to fight your scores.
What the engine actually shows you
A directional band, a confidence read, and the dimensions that lifted or dragged. Use the explanation to check whether the level holds up under calibration.
When the wizard finishes, you'll see a recommended , a confidence rating, a possible boundary, and any sanity flags. Read all four. The by itself is the smallest part of what the tool produces.
| Field | What it tells you |
|---|---|
| Level + name | Where the math landed (e.g., P4 Advanced) |
| Weighted score | The number behind the level |
| Confidence | How tightly the dimensions agree |
| Boundary | Whether the role is sitting near an adjacent level |
| Move-up / Move-down | Which dimensions would shift the level |
| Flags | Inconsistencies worth a second look |
| Calibration prompt | A question to take into the calibration session |
Reading a leveling result
- Field
- Level + name
- What it tells you
- Where the math landed (e.g., P4 Advanced)
- Field
- Weighted score
- What it tells you
- The number behind the level
- Field
- Confidence
- What it tells you
- How tightly the dimensions agree
- Field
- Boundary
- What it tells you
- Whether the role is sitting near an adjacent level
- Field
- Move-up / Move-down
- What it tells you
- Which dimensions would shift the level
- Field
- Flags
- What it tells you
- Inconsistencies worth a second look
- Field
- Calibration prompt
- What it tells you
- A question to take into the calibration session