From calibration to comp: connecting year-end ratings to merit increases

Calibration takes hours of leadership time every cycle. Leaders compare ratings across managers, teams and departments so that a 4 in Engineering means the same thing as a 4 in Sales. That work feeds three decisions: the performance review (and the ratings that the employees see about themselves), who gets promoted (or doesn’t) and what each person's compensation looks like after the cycle is over.

The third decision is where the work most often gets lost. In many companies, ratings are finalized in one process and pay is decided in another, weeks later, in a different tool. The calibrated rating becomes one input among several, and each manager falls back on their own judgment about who deserves what. The consistency calibration created never reaches pay.

Here's why ratings and pay need to be connected, which ratings should inform pay and how to turn a calibrated rating into a compensation decision you can explain and defend.

What calibration is for

Calibration creates consistency across an organization, irrespective of individual manager preferences, styles and department cultures. It normalizes ratings for the performance review itself. It also informs who gets promoted and what their compensation increase looks like.

That last part only works if the rating measures the right thing. Our recommendation: base the ratings that inform compensation on actual, demonstrated performance, and leave potential out of pay decisions. Potential can carry tremendous bias, and it's a metric for succession planning. It tells you who to develop. Pay should reward the competencies and results someone has already shown.

For pay-for-performance cultures, this is the whole point. Calibrated ratings are how you show that pay matches, and is justified by, actual output.

Before you set a percentage, check the starting point

A merit matrix takes a rating and a compa-ratio and returns a percentage increase. It's simple, which is why most companies use one. The catch is that driving toward a percentage assumes the person's compensation is in the right spot to begin with. If the base is off, the increase carries the error forward.

A few signs the starting point is off:

• The same people sit below target every year. The matrix gives them a slightly bigger percentage each cycle, and the gap never closes.
• Your strongest performers get increases close to everyone else's. A point or two of spread, applied to different bases, is a thin version of pay for performance.
• Out-of-guideline exceptions keep growing. Managers are working around the matrix to do what it should have done.
• New hires land above tenured high performers. The market moved, and percentage increases didn't keep up.
• Managers can explain the percentage but not the pay. "You got 4% because you're a 4" doesn't answer "am I paid fairly for what I do?"

If none of these sound familiar, your matrix is doing its job. If several do, look at where people sit in their bands before you decide on a percentage.

Start with where each person should sit in their band

Two factors drive where someone belongs in their salary band:

• Competency progression. How far along is this person in the skills and scope of the role? Someone new to the level sits lower in the band. Someone working at the top of it sits higher. Measure this by demonstrated competency. Tenure alone doesn't move someone up the band.
• Performance. What is this person delivering, based on their calibrated rating?

The two may seem the same, but they're different. Someone can be on the lower end of the band and contribute at a higher level than a colleague who looks more senior but is having less of an impact. Pay should reflect both.

Put the two together and every employee has a target position in their band. The merit increase is the distance between where they're paid today and where they should be. For one person that's 9%. For another it's 1%, because they're already there.

You can still publish a matrix so managers have guidelines to work from. It becomes an output of this work, built from your real bands and your real people.

Model the budget bottoms-up

Budgets are usually set top-down, as a percentage of payroll. Bottoms-up modeling lets you spend what performance and the market actually dictate.

1. Confirm your salary bands are current and every employee is mapped to one.
2. Set each person's target band position from their competency progression and calibrated rating.
3. Calculate the gap between current pay and target for every employee.
4. Add up the gaps. That total is your ask.
5. Bring it to Finance with the detail by team and level.

Team budgets come out of the model. A team of underpaid strong performers gets a larger share because the data says they need it. Finance should be in the room for this, working from the same headcount and salary data as HR.

When the need is bigger than the budget

Sometimes the model says 5.2% and the company can fund 4%. When that happens, the comp model you're adhering to decides what gets funded first. Common models include:

• Gap and performance first. Fund the largest gaps for the strongest performers, at any level.
• Protect the lower bands. Hold executive increases to fund corrections for more junior team members. Executives carry more of their pay in bonus and equity, and one deferred increase can fund several adjustments.
• Retention and market pressure. Fund the roles that are hardest to replace or have moved most in the market.
• Change the timing or the vehicle. Close the gap over two cycles, or use bonus or equity where base can't stretch.

Pay equity gaps come first under any of these.

Whichever model you choose, choose it before anyone sees the names. A rule picked after the model shows who loses reads as favoritism. It's the same principle as calibration: consistency that doesn't depend on who's in the room.

Protect the plan while managers make their calls

A bottoms-up model gets you a funded plan. Guardrails during the cycle keep you there.

Lock the rating. Carry the calibrated rating into the comp cycle automatically and lock it once calibration closes. If managers can quietly adjust pay without reference to it, the rating stops meaning anything. Next year, everyone knows calibration is theater.

Flag exceptions for review. Hard stops frustrate managers and push decisions offline. Flag any recommendation that moves someone away from their target position, or any team that goes over its allocation, and route it to the next approver with the manager's justification attached.

Hold a small reserve. Keep 5% to 10% of the pool at the HR or executive level for promotions, equity adjustments and retention cases that come up during review.

Watch for pay equity drift. Before final approval, check whether recommended increases widen any gaps across gender, race or other protected groups at the same level and rating. It's far easier to fix in the cycle than to explain afterward.

Give managers the context to explain the pay

The comp conversation is where employees decide whether the process was fair. Most managers walk into it with a number and very little else. Give them what they need to explain it.

In the planning view, each manager should see, for every direct report:

• Calibrated rating and any notes from the calibration session
• Current salary, band and position in band
• Target position, and the gap between the two
• Last increase amount and date
• Goal progress and key review feedback

For the conversation itself, equip managers with a short script:

1. Start with the rating and what drove it, using specifics from the review.
2. Show the band, where the employee sits in it today and where their competencies and performance say they should sit.
3. State the new salary and the effective date.
4. Cover what's next: what it takes to move up in the band or reach the next level, and how pay grows with it.

An employee who gets a smaller increase than a peer will still feel the gap. When the manager can show the band, the position and the path forward, the conversation shifts from "why did I get less?" to "what do I need to do next?"

After the cycle, generate total rewards statements so employees can see their full picture: base, bonus, equity and benefits.

Checklist: connecting calibration to comp this cycle

1. Calibrate on demonstrated performance, and keep potential for succession planning.
2. Lock calibrated ratings before the comp cycle opens.
3. Confirm salary bands are current and every employee is mapped to one.
4. Set a target band position for each employee from competency progression and performance.
5. Model the total need bottoms-up and review it with Finance.
6. Agree on your funding philosophy before anyone sees names.
7. Set the reserve and the approval path for exceptions.
8. Load ratings, bands, target positions and budgets into the manager view.
9. Run a pay equity check before final approval.
10. Brief managers on the conversation structure before they meet with employees.
11. After the cycle, compare outcomes to targets and adjust for next year.

Close the loop with one system

Calibration and merit reviews are two halves of the same decision. Modeling pay from the bottom up takes ratings, bands, current pay and budget in one place. In separate tools, that means exports, pasted spreadsheets and version conflicts reconciled by hand.

That's how ChartHop works. Performance and Compensation share one data layer, so calibrated ratings flow directly into comp cycles alongside band data, budget and review history. Managers see guidelines and remaining budget as they plan. HR configures approval workflows and flags outliers centrally. And ChartHop AI can model the budget impact of proposed changes and surface pay equity gaps before anything goes final.

See how ChartHop connects performance and pay. Book a demo →