How this tool produces an answer, and where it stops.

What this tool is for (and isn't)

This is a directional education tool for HRBPs, managers, and Total Rewards leaders who want a credible way to teach job architecture and prepare better leveling intake. It is not a final pay or compensation decision engine, and it does not reproduce proprietary Radford, Mercer, or WTW level descriptors, point-factor methods, survey job descriptions, or job codes.

Privacy and what to keep out

Keep the following out of any input field, downloaded summary, or screenshot you share from this site:

• Employee names or other identifiers.
• Compensation details such as base salary, bonus, equity, or target incentive.
• Protected-class information including age, race, gender, religion, disability, parental status, or veteran status.
• Confidential business plans such as upcoming reorganizations, unannounced layoffs, or pending acquisitions.

What the tool does with your inputs:

• Scores stay in your browser. Refreshing the page clears the wizard. Closing the tab clears the calibration session.
• The site does not log inputs to a server, write user data to local storage, or send anything to a third party. The only saved preference is your dark or light theme choice.
• Downloaded text summaries are generated in the browser and saved to your device. The site never sees them.

Product behavior is a starting point, not a privacy guarantee. Treat anything sensitive as off-limits even if the tool would not retain it.

Scoring model

The wizard asks about six core dimensions of a role: scope, complexity, autonomy, influence, knowledge, and business impact. Each dimension has a 7-point scale.

The engine averages the six. Manager-track roles add a seventh dimension (people leadership) at 20% weight. Executive-track roles add a seventh (strategy ownership) at 30% weight. The weighted score maps to a level on the IC, manager, or executive ladder. Bands are documented in the engine reference.

Confidence rating

Confidence is the standard deviation of your dimension scores. If your scores agree closely, confidence is high. If they're scattered, confidence is low. Low confidence usually means the role is a hybrid, a stretch, or a boundary case worth talking about in calibration.

Boundary detection

If the weighted score sits within 0.2 of an adjacent level threshold, the result flags the boundary (e.g., "P4 (boundary P3/P4)"). The level you see is the band the score actually falls in. The boundary tells you that small shifts in evidence could move it.

Sanity flags

Flags are informational. They don't block the result. They surface when something in the inputs looks inconsistent: an autonomy score far higher than complexity, a people-leadership score with a narrow scope, an executive selection with low strategy ownership, or scores that vary widely across dimensions. Each flag suggests something worth a second look in calibration.

Sources

1. [1]Aon Job Architecture

   Why I cite it

   Public Aon material frames job architecture around functions, families, jobs, career levels, and titles, and connects it to compensation, career frameworks, pay equity, and pay transparency.

2. [2]Radford McLagan Compensation Database reporting features

   Why I cite it

   Aon describes the database as a major source of compensation data and tools for benchmarking, peer groups, job matching, employee-vs-market analysis, and executive regression.

3. [3]Radford McLagan Job Offers Data

   Why I cite it

   Aon states that Job Offers Data and RMCD data use a consistent job architecture and leveling structure for apples-to-apples comparisons.

4. [4]Aon Compensation 101

   Why I cite it

   Aon explains why job levels, job families, survey participation, and job matching are critical Total Rewards skills.

Limitations

Limitations:

1. The dimensions are a working model. Your company's real architecture may use four dimensions, or eight, or different ones entirely.
2. The level bands reflect common-market shapes. P4 here is a directional pointer, and your company's P4 may sit somewhere else on the same shape.
3. The engine never sees your company's salary structure. Pay range design is a different decision and requires licensed market data.
4. Sanity flags catch obvious inconsistencies. They miss subtle ones. Calibration still beats automation for boundary cases.

Radford / Aon-friendly usage

Use this tool to start the conversation. For the actual job matching, market pricing, peer-group construction, and pay range design that production Total Rewards work requires, you'll want licensed Radford or Aon resources (or your survey provider of choice). The tool intentionally does not reproduce proprietary level descriptors, point-factor methods, or survey job content. The public Aon and Radford framing this site references is listed under Sources.

Disclaimer

This tool provides directional job architecture education and a directional level recommendation based on general market practices. It is intended for guidance only and should be used alongside internal calibration, licensed market data, and professional judgment. It does not reproduce proprietary Radford, Mercer, WTW, or other survey-provider methodology, and it is not a compensation, legal, compliance, or final decision engine.