Most compensation cycles are working on two problems at the same time. Inflation has been eating into real wages, and you still need to recognize the people who actually earned it this year. The typical answer is folding both into a single raise, and that raise tends to do a poor job of either.
| Merit Increase | Cost of Living Adjustment | |
|---|---|---|
| Purpose | Recognize individual performance | Protect purchasing power against inflation |
| Who receives it | Varies by individual | All eligible employees at a uniform rate |
| Calculation basis | Performance rating and compa-ratio | Regional CPI data or company-set floor |
| Discretionary? | Yes | No |
| Pay range interaction | Constrained by band position | Unaffected by band position |
| What it tells employees | “Your contribution earned this” | “The economy required this” |
What Each Mechanism Is Actually For
Merit is supposed to mean something. When everyone on your team lands between 3 and 5 percent regardless of what they actually produced, the label stops working. Your strongest performers notice when the gap isn’t wide enough to reflect what they put in.
COLAs aren’t a reward, they’re a correction. Inflation moved, and the organization is making sure real wages don’t quietly fall behind. Every eligible employee at a given location gets the same rate, because the economy didn’t perform differently for different people. The moment you start varying it by individual, you’re no longer running a COLA.
A 4.5% raise that was supposed to cover a 2.5% COLA and a 2% merit increase ends up saying nothing about either. The employee hears the number, compares it to what a colleague got, and draws their own conclusions.
Those conclusions tend to be wrong, and the employees most likely to get it wrong are often the ones you’d most like to keep.
The Three Failure Modes
Each of the following patterns is common enough that most compensation teams have seen at least one play out. In every case, one mechanism ends up doing two different jobs.
Blending the Budget and Calling It Merit
When CPI peaked at 9.1% in June 2022, many organizations quietly inflated their merit budgets to compensate for inflation rather than formally adding a COLA line. A company with a historical merit budget of around 3.5%, which is the median reported in WorldatWork’s 2024 Salary Budget Survey, stretched it to 6% and told managers to distribute based on performance.
What managers actually did was more predictable. Most gave everyone at least 4% to avoid the conversation about real wage cuts, which left only 2 percentage points of differentiated budget for genuine merit recognition.
High performers received 6% while everyone else landed between 4 and 4.5%, a spread too narrow to mean anything as recognition. Total spend nearly doubled what a properly structured cycle would have cost.
Everyone left the process frustrated for different reasons. High performers felt the recognition was thin. Average performers couldn’t figure out why they’d gotten less than a colleague. Finance had no clean story for the overage.
Using COLA to Avoid Having a Merit Conversation
Some organizations go the other direction. A flat 3% COLA during a year when inflation ran at, say, 2.5% sounds responsible, right up until you consider what it communicates to your top performer. They just received the same increase as the employee who has been quietly underperforming for 18 months.
The organization spent money and told its workforce that performance doesn’t affect what people earn. That’s a less expensive mistake in dollars than blending the budget. It tends to be a worse one in terms of who decides to leave.
The Compa-Ratio Problem
Merit increases interact with pay ranges in a way COLAs don’t. An employee at 118% of their range midpoint shouldn’t receive the same merit increase as someone at 85%, even if their performance is equivalent. The first is already paid above market; another 4% makes that worse without any market logic behind it.
COLAs don’t carry that constraint. They reflect external economic pressure regardless of where someone sits in the range. When the two mechanisms get merged, the compa-ratio discipline that should govern merit gets buried under inflation logic, and neither one works the way it should.
How to Run Both in the Same Cycle
The fix for all three patterns above is the same. Run merit and COLA as distinct mechanisms with separate budgets, and separate them before the planning cycle opens — not after the damage is done.
This assumes an organization with a dedicated compensation function and roughly 150 or more employees, the threshold where separate planning tracks are operationally realistic. Smaller teams can often work through both in a single finance conversation, even when the components get tracked separately in payroll.
For organizations above that threshold, three distinct line items make the structure work:
- COLA budget: a flat percentage applied to all eligible employees, based on regional CPI data or a company-determined inflation floor. Not discretionary. Every employee at a given location gets the same rate.
- Merit budget: a pool distributed by performance rating and compa-ratio, calculated independently from the COLA. A manager whose team receives a 2.5% COLA still has their full merit pool to allocate. The COLA doesn’t reduce it.
- Combined effective increase: what the employee actually receives. The sum of both components, tracked separately through planning and communicated separately to the employee.
In CompLogix, these map to separate compensation components within the same planning cycle. Managers see their COLA allocation and their merit pool as distinct line items, which means the COLA is never available to redistribute.
At the reporting stage, merit spend pulls by performance rating and COLA spend by location, without needing to reconstruct anything from a merged field. That matters when you’re auditing equity or explaining a variance to finance.
Planning principle: The COLA is already allocated before managers open their planning workflows. It’s the floor, not part of what they’re distributing. Merit decisions should be made as if the COLA doesn’t exist — because in planning terms, it doesn’t.
The Communication Is Half the Work
Getting the structure right is only half the work. The other half is making sure managers can actually explain it to employees.
A manager who received a merged number can’t do that, because they’ve never seen the components separately. The conversation that works looks something like this:
“Your salary is increasing from $84,000 to $87,500, effective March 1. Of that $3,500, $2,100 is a cost of living adjustment we’re applying to everyone at your location. The remaining $1,400 is a merit increase based on your performance — specifically your contributions to the product launch in Q3 and how you handled the team through the restructuring.”
That conversation gives the merit increase its meaning. It also gives the employee honest information about what the COLA component is, which matters more than you might expect during years when people are paying attention to their purchasing power.
CompLogix’s planning workflow keeps both components visible as separate line items, so managers walk into that conversation already knowing both numbers.
What to Do When You Can’t Afford Both
Pick one and say which one you picked.
A COLA-only year tells employees that inflation is being addressed for everyone but the merit budget is constrained this cycle. A strong performer who hears that explanation will process the COLA very differently than one who receives the same dollars relabeled as merit recognition. One of those conversations is honest.
A merit-only year tells employees you’re recognizing performance but not offsetting inflation. High performers see real gains. Average performers may see their purchasing power flat or slightly negative, and they’ll do that math on their own. That’s a defensible position — but only if you’ve had the conversation rather than hoping nobody notices.
The harder conversation is usually the internal one. When leadership wants to give everyone a flat raise but call it performance-based, that isn’t a communication choice. It’s a data integrity problem that surfaces the first time anyone tries to correlate merit spend with retention data and the numbers don’t line up.
Handling Employees Already at the Top of the Range
Employees above their range maximum are where the distinction between merit and COLA becomes impossible to sidestep.
A merit increase that would push someone above their range maximum runs into a real boundary. You’re paying above market, which requires either a formal exception or a range adjustment. That’s a legitimate constraint.
A COLA for the same person sits outside that logic entirely. If inflation ran 2.5% and the organization isn’t adjusting this person’s pay, it’s cutting their real compensation on the basis of where they sit in a band — not for any economic reason. That’s a deliberate decision, and it deserves an explicit conversation rather than being silently absorbed into range mechanics.
One option is to pay the COLA component as a lump sum for above-range employees rather than adding it to base salary. An employee at $120,000 and 115% compa-ratio who receives a 2.5% COLA as base salary ends up with a $3,000 permanent increase that pushes the above-market problem further.
As a lump sum, they get the same $3,000 without that consequence. This works when the employee already understands they’re at range ceiling. If that conversation hasn’t happened yet, that’s where to start.
If this cycle’s planning hasn’t separated the budget lines yet, that’s the first thing to fix. The manager conversations, the reporting, and the year-over-year analysis all follow from that structure. None of it works when the two mechanisms are merged from the start.
Frequently Asked Questions
Can you give a merit increase without a COLA?
Yes, and it’s sometimes the right call. A merit-only cycle rewards performance without offsetting inflation, so high performers see real gains while average performers may see purchasing power flat. The retention risk sits with strong middle performers. Clear communication about why COLA was skipped is the main management tool.
Should merit increases and COLAs go into effect at the same time?
Generally yes. A single payroll change event is simpler to administer, and employees experience both components together. What matters is that the planning, approval, and communication treat them as distinct. There’s no analytical benefit to staggering the effective dates, and doing so typically creates more confusion than clarity.
How do you handle COLA for employees in different geographic locations?
COLA rates should vary by location, not by individual. The Bureau of Labor Statistics publishes regional CPI data by metro area, and the standard approach is setting a rate per location and applying it uniformly. For fully remote employees, most organizations use the employee’s home address as the anchor.
What if our salary ranges need updating during the same cycle?
Finalize range adjustments before merit planning begins. If ranges shift after managers submit recommendations, the compa-ratio logic behind them is no longer valid. COLA can be calculated at any point since it doesn’t depend on range position, but sequencing everything after ranges are set avoids a second round of approvals.
What do you do when union contracts specify COLA but merit is discretionary for salaried employees?
When hourly employees receive COLA under a collective bargaining agreement, salaried employees typically expect the same. Skipping it creates an equity problem that shows up as questions from your strongest performers. Where the two groups legitimately diverge is on merit. Union merit is governed by contract; salaried merit stays discretionary.