Merit Increase Meaning: Definition, Examples, and Guide

Sooner or later, someone on your team will ask why a colleague got a bigger raise. At a small business without an HR team, you are the one who has to answer, and this guide is about making sure your answer holds up.

The first merit cycle I ran was not a cycle. It was eight separate conversations, each one held in isolation, each one ending with a number I made up on the spot based on how the conversation had gone. Two people got 5 percent. One got 2 percent because he had not asked for more. The rest landed somewhere in between for reasons I could not have defended in writing.

Nine months later, two of those people found out what the others got. The 2 percent was not the problem. The problem was that I had no answer to the question "why?" that did not amount to "because that is what came out of the conversation." That is not a compensation philosophy. That is a series of negotiations, and the people who negotiate hardest win them.

A merit increase is supposed to be the opposite of that. Below, I cover what it means and how it differs from a raise, a cost-of-living adjustment (COLA), a promotion, and a bonus. Then come the current average, a merit matrix that turns eight arguments into eight lookups, the budget math, and a full cycle a small team can run.

What Is a Merit Increase?

A merit increase is a permanent raise to an employee's base salary awarded because of their performance. The trigger is how well the person did their job, not the cost of living, not how long they have been there, and not a change in what the job is.

Three terms in that definition do the heavy lifting. Permanent separates it from a bonus. Performance separates it from a cost-of-living adjustment. Criteria set by the employer separates it from a legal entitlement, which is what makes documentation matter so much: the criteria are yours, so you have to be able to show you applied them the same way to everyone.

The permanence is the part small business owners consistently underestimate. A 4 percent merit increase is not a 4 percent cost this year. It is a 4 percent increase to that person's payroll cost forever, and next year's percentage increase is calculated on the new, higher base. Grant 4 percent five years running and the salary is 22 percent higher than where it started, not 20.

Merit Increase vs Raise: What Is the Difference?

A merit increase is a type of raise. Raise is the umbrella term for any increase to pay. Merit increase is the specific variety awarded for performance. Every merit increase is a raise; most raises are not merit increases.

This sounds like a semantic distinction, but it is not: the two words carry different obligations. If an employee asks for "a raise," the conversation could be about market rate, cost of living, scope creep, or performance, and each of those has a different answer. If an employee asks about their "merit increase," they are asking a narrower question: what did my performance earn me, and how did you decide?

The practical consequence is that you need to know which conversation you are in. Someone who is underpaid relative to the market does not need a merit increase; they need a market adjustment. Giving them 3 percent for performance does not fix a 15 percent market gap.

Someone whose job has quietly doubled in scope needs a promotion, not a merit increase. Using merit as the answer to every pay question is how companies end up with people who are excellent, well-reviewed, and 20 percent underpaid.

Merit Increase vs COLA vs Promotion vs Bonus

Four kinds of pay increase get confused with each other constantly, and the confusion is expensive because each one solves a different problem. Here is what separates them.

A cost-of-living adjustment goes to everyone and tracks inflation. It says nothing about performance. A promotion changes the job. A bonus is a one-time payment that leaves base pay untouched, which is why it is the cheapest way to reward a one-off result and the worst way to fix a salary that is too low.

A fifth instrument belongs next to these four: the market adjustment, which corrects pay that has fallen behind the market for the role. Here is when to reach for each one.

Most small businesses use exactly one of these and try to make it do the work of all five. The result is either merit increases that quietly function as COLAs, or bonuses substituting for salaries that should have been raised two years ago. Naming the instrument correctly is most of the discipline.

Merit Pay and Merit Pay Increases

Merit pay is the system. A merit increase is the output of that system for one person in one cycle. When people say "merit pay increase," they usually mean the same thing as a merit increase, but the phrase is worth unpacking because it points at something larger than a single number.

A merit pay system requires four things to exist before the first dollar moves: written performance criteria, a rating scale that someone actually applies, a budget, and a rule connecting the rating to the money. Skip any one of those and you do not have merit pay. You have discretionary raises with a nicer name, which is precisely what I had in my first cycle.

The Department of Labor's framing is useful here because it is deliberately minimal: pay-for-performance is a raise based on criteria the employer sets. It says nothing about what those criteria should be. That is your job.

What Is a Typical Merit Increase Percentage?

US employers projected an average merit increase budget of 3.2 percent for 2026, with total salary increase budgets of 3.5 percent covering merit, promotions, cost-of-living, and other adjustments combined. Actual increases came in slightly below projection.

Two things to take from that. First, budgets have been flat for three years, which means the elevated pandemic-era increases are over and you should not benchmark against them. Second, the 2026 merit projection overshot the actual by a tenth of a point (3.2 percent planned, 3.1 percent paid), so if you build a budget at exactly the projected number you are likely already above the market.

The flatness is easier to see across the whole series. Earlier compensation planning surveys put actual merit increases at 3.3 percent in 2024, with employers planning 3.3 percent again for the year after. A drift from 3.3 to 3.1 percent across two cycles is the entire movement, so anyone waiting for the number to snap back is waiting on nothing.

The 2027 planning cycle looks the same. SHRM reported in August 2026 that a global survey of more than 34,000 companies, 1,650 of them in the US, puts US salary increase budgets at an average of 3.4 percent in 2027.

The same survey recorded an actual 2026 budget of 3.5 percent, so the decline is slight. For a small business that means planning the next cycle against the same band you used for this one.

The number that matters more than the average is the spread. A 3.2 percent budget does not mean everyone gets 3.2 percent. It means the average across the team is 3.2 percent, which requires strong performers to get 5 or 6 percent and weak performers to get 0 or 1 percent. A company where everyone gets exactly the budget number has a payroll adjustment, not a merit system.

How the Merit Matrix Works

A merit matrix is a grid that crosses performance rating against position in the salary range and returns the increase percentage. It exists to make the decision a lookup rather than a negotiation, which is what stops the loudest person from getting the biggest number.

The second dimension is what makes this a matrix rather than a list, and it is the part most small companies leave out. Two people can earn the same rating and deserve different percentages, because one of them is at the bottom of the salary band and the other is at the top.

Someone at the bottom has room to move. Someone at the top of the band who is still exceeding expectations does not need a bigger merit increase. They need a promotion conversation, because they have outgrown their range.

The measure for the second dimension is compa-ratio: current salary divided by the midpoint of the salary range for that role. A compa-ratio of 0.85 means someone is paid 15 percent below the midpoint. A compa-ratio of 1.15 means they are 15 percent above it. If you do not have salary ranges yet, build them before the matrix, because without a range there is no position to measure.

How to Calculate a Merit Increase

New salary equals current salary multiplied by (1 plus the increase rate). That is the entire formula.

An employee earning $62,000 who receives a 4 percent merit increase moves to $62,000 times 1.04, which is $64,480. The annual increase is $2,480. Divided across 26 biweekly paychecks, that is about $95 per paycheck before tax.

What a 2, 3, or 5 Percent Merit Increase Is Worth

On a $60,000 salary, a 2 percent merit increase is $1,200 a year, or about $46 per biweekly paycheck. The same salary at 5 percent is $3,000. Running the whole ladder against one salary is the fastest way to see what the percentage you are about to grant actually says to the person receiving it.

A 2 percent increase is the one that most often needs explaining. Against a 3.2 percent average it reads as below average, and the employee will hear it that way whether or not you say so. If performance is the reason, say that. If the reason is that they already sit above their range midpoint, say that instead.

The math only gets harder at the budget check, where every individual increase has to fit inside one fixed pool.

How to Set a Merit Increase Budget

Multiply total payroll by your merit budget percentage. That is the pool. Fix it before you look at a single person, because a pool decided after the fact is not a pool.

Notice what makes this work: the differentiation is real. Two people get 5.5 percent, which is meaningfully above market. One person gets nothing. The average across the team is exactly 3.2 percent. If instead you gave all ten people 3.2 percent, you would spend the same $20,800 and communicate nothing at all about performance.

The uncomfortable part is the zero. A merit system where nobody ever gets zero is a system where the top performers are subsidizing the bottom, because the pool is fixed. If you are not willing to give anyone zero, you do not have a budget problem. You have a performance management problem.

Running Your First Merit Cycle

The sequence matters more than the sophistication. A simple process run in the right order beats a sophisticated one run backward, and the most common way to run it backward is to decide the numbers first and reverse-engineer the ratings to fit.

The workbook below is those seven steps in the order they have to happen. Fill the budget tab first with no names in front of you, set your own percentages in the matrix tab before anyone is rated, then work the employee tab left to right so the rating and its evidence are written down before the salary column is touched. The percentages in the matrix are deliberately blank: they are your decision, and they should be made once for the whole cycle rather than person by person.

Start documenting this before you think you need to. Compensation decisions made informally on a small team become impossible to reconstruct once the team has doubled, and the first person to ask a hard question about them will be the one you least want to lose.

How to Communicate a Merit Increase

Put it in writing, state the specific contribution that earned it, and confirm the increase is permanent. Those three things answer the questions employees actually have, and answering them in the letter prevents the follow-up meeting where they ask anyway.

Two lines in the letter carry most of the weight. One names the contribution, and it should name the outcome rather than the effort. The other says plainly that this is a permanent change to base salary, not a one-time payment.

The conversation that follows the letter should be short. The employee already has the numbers, so your job is to explain the reasoning and answer the one question they always ask, which is some version of "what would have made it bigger?" Have an actual answer to that. "Keep doing what you are doing" is not an answer; it is a way of avoiding one.

Deliver the news of a smaller-than-expected increase yourself, in person, and do not blame the budget for a decision you made. If someone got 1.5 percent because their performance was mixed, tell them that. You made the call, not the budget, and pretending otherwise costs you the one thing that makes the next cycle work: their belief that the ratings mean something.

Pros and Cons of Merit Increases

The strongest argument against merit increases is not that the concept is wrong. It is that most companies implement it badly, and a badly implemented merit system is worse than no system at all, because it puts a veneer of objectivity on decisions that were made on instinct. That is the failure mode worth understanding in detail.

Bias, Pay Compression, and the Compounding Problem

Merit increases compound, which means any bias in the ratings compounds too. This is the mechanism by which a small, defensible-looking difference becomes an indefensible one over a few years, and nobody notices it happening because each individual decision looked fine.

Two people start at $60,000 in the same role. One consistently receives 4 percent, the other consistently receives 2.5 percent. After five years the first earns roughly $73,000 and the second roughly $67,900. The gap is over $5,100 and growing, and it was never the product of a single big decision. It was five small ones, each of which felt reasonable at the time.

The other structural problem is pay compression: new hires arrive at market rate, existing employees have only ever received 3 percent merit increases, and the market moves faster than 3 percent. After a few years of this, the person you hired last month earns more than the person who trained them. Merit increases alone will never fix this, because compression is a range problem and merit operates within the range.

Both problems are why the pay equity check belongs in the cycle rather than after it. Under the Equal Pay Act, a pay difference between employees doing substantially equal work is lawful only where it rests on a seniority system, a merit system, a system measuring earnings by quantity or quality of production, or a factor other than sex.

Those four exceptions are affirmative defenses, which means the employer is the one who has to prove them. The statutory text says nothing about documentation, but EEOC guidance expects a merit system to use predetermined criteria, be communicated to employees, and be applied consistently. An undocumented merit system does not meet that bar.

The Case Against Flat Increases

Giving everyone the same percentage is tempting because it is fast, it looks fair, and it eliminates every difficult conversation in one move. It also eliminates the entire point of a merit system.

You may have seen coverage of a shift toward uniform across-the-board raises, sometimes described as spreading the budget evenly rather than differentiating. The data does not support the idea that this is where the market has landed.

The follow-up survey of 756 employers found only 4 percent gave equal across-the-board increases. The majority still set increases on some combination of individual performance, position against market value, or position against peers. Differentiation is still the norm by a wide margin.

What the October planning survey does show is a related pattern worth watching: 83 percent of employers said they would distribute the budget equally across the organization rather than directing more of it toward high-demand skills or critical market gaps. That is a different failure. It is not flat pay for individuals; it is flat allocation across teams, which quietly underfunds the roles you most need to keep.

The argument against flat increases is straightforward. If a strong performer and a weak performer both receive 3.2 percent, you have told the strong performer that their performance was worth nothing, and you have told the weak performer that theirs was fine. You spent the entire budget and sent no message with it. The people most likely to leave after a flat cycle are the ones who most deserved more.

Merit Increases Without a Dedicated HR Team

You do not need a compensation consultant or a survey subscription to run a defensible merit cycle at a small business. You need four artifacts, and all four fit in a spreadsheet.

The harder problem is not building those four things. It is that at a small company, the performance history, the salary data, the offer letter, and the pay change all live in four different places, and reconstructing the decision six months later means opening four systems and an email thread.

That is part of the gap I built FirstHR to close. Employee profiles keep each person's record and documents in one place, and task workflows keep the cycle on a calendar instead of on whoever happens to remember it.

Document management with built-in e-signature closes the other half. The increase letter ends up stored where the employee record lives rather than in a sent folder. The math is not the hard part of a merit cycle. Reconstructing your own reasoning a year later is.

Common Merit Increase Mistakes

Each of the seven mistakes below undoes part of the process above, and each one has a straightforward fix.

Every one of these is a variation on the same underlying error: treating merit as a number to be arrived at rather than a conclusion to be reached. The number is the last thing that should happen in a merit cycle. In my first one, it was the first.

Frequently Asked Questions