A merit cycle is always trying to do several things at once.
Reward performance. Retain the people you can't afford to lose. Correct market drift before it turns into an attrition problem. Promote those ready for the next step.
But when the budget is tight, those priorities pull in different directions, and something has to give.
It’s a gap most process guides skip over: the difference between how a merit cycle is designed on paper and how it actually runs once real managers, real budgets, and real employees are involved.
So in this guide, we’ll be filling that gap – sharing what a solid merit cycle process looks like, as well as advice and honest learnings from Ekaterina Potter, Global Director of Reward and Belonging at commercetools, and Daniela Černá, Global Compensation and Benefits Lead at Aeven on the parts that break in practice, and how to change it for next time.
A merit cycle is a structured process, usually run annually or biannually, that rewards employees for their past year's performance.
It may also include pay adjustments for market corrections, pay equity fixes, and promotions too – in which case it might, more broadly, be referred to as a compensation review.
There's no universal template for a merit cycle – how broad the scope is, how tight the budget, how much discretion managers get, all depend on the company.
But there is a reliable order of operations, from deciding what the cycle actually covers through to reviewing how it went.
That's what the eight steps below cover, complete with plenty of advice and lessons learnt from both Ekaterina and Daniela.
Before anything else, decide what this merit cycle is actually for.
A pure merit cycle might reward performance only.
A broader compensation review, which is what most companies actually run, also covers market adjustments, pay equity fixes, and sometimes promotions.
This is a Reward-and-leadership decision, and it needs a foundation already in place to work from: job architecture, market-anchored salary ranges, and a working performance management process.
If those don't exist yet, tackle those first – trying to run a merit cycle without them is what Ekaterina describes as her starting point at commercetools, where "our compensation review, before I joined the company, was more like a profit-sharing moment in the past."
The scope of a merit cycle is broad at commercetools – basically all adjustments except promotions sit within the cycle. "It's not only merit that we reward during this process," Ekaterina says. "Merit and rewarding performance is a large part of it – but also it's about market adjustments and addressing any internal gaps.”
Because of this, commercetools' cycle starts with an analysis of where those gaps actually are months before the review actually starts.
To give one example of what this analysis informs, tenure is always a big culprit.
"This is a classic problem in the tech industry, where you have people who are the strongest performers, know our product the best, have been here for a while," she says.
"If you bring in a newcomer, they might be brilliant, but they still need to learn about the product – and those newcomers come in at a compensation that the long-tenured, very wonderful employee needs years to catch up with.”
“So we preemptively look at these gaps, crunch the data, and then make sure we allocate budget into those teams."
At commercetools, promotions run on a separate budget and can happen at any point in the year, rather than being tied to the cycle. “We have an ownership-driven culture, so we really value that managers are trusted to promote their employees at any time," Ekaterina says.
But what works for one company might not work for another, so this first step is about aligning your merit cycle with your compensation philosophy, before anything else kicks off.
"It's not only merit that we reward during this process – it's about market adjustments and addressing any internal gaps.”
Ekaterina Potter
Global Director of Reward and Belonging at commercetools
Once you have your scope agreed, it’s also important to set clear rules on employee eligibility – rather than leaving it to case-by-case judgement when budget is already tight.
At Aeven, for instance, Daniela's team excludes anyone hired, or given a salary increase, in the last three months.
Recent hires and recent increases are excluded on a straightforward logic: someone four months into the role hasn't been through a full performance cycle yet, so there's no fair basis to rate them against colleagues who have – and their pay was set against current terms too recently for there to be any drift to correct.
"If someone was hired in January and the salary cycle happens in March, they're not eligible for any salary increase in this cycle,” she says. "It’s for fairness, but this mechanism also helps us have some extra budget to cover those other gaps."
With scope decided, it’s time to negotiate the actual number – and this should happen well in advance of the cycle starting.
Commercetools launches its merit cycle in February, but the budgeting conversation with Finance starts the previous October. "We start working on our merit cycle way in advance," Ekaterina says. "It starts from October, through the budgeting process that concerns the whole company."
"It is always all about multiple demands on a finite budget," she explains.
"Finance brings us what they think the company can afford, and we bring into the conversation what we think we need to do for us to stay competitive, to close the gaps, to make sure that we have long-tenured employees rewarded properly, and that we don't have any gender pay gaps."
"It is always all about multiple demands on a finite budget."
Ekaterina Potter
Global Director of Reward and Belonging at commercetools
With scope and budget locked, decide the mechanism that turns a performance rating into an actual percentage increase, and who gets to make that call for each employee.
The simplest version is a flat percentage per rating: everyone who exceeds expectations gets the same increase, regardless of anything else.
That’s easy to run, but it ignores the context of where someone already sits against the market and their peers – which means your best-paid performers might keep pulling further ahead of equally strong performers who joined more recently or negotiated less hard.
Our view is that band position should factor into the increase even in a pure merit cycle – a flat percentage rewards performance but does nothing to correct the drift that performance alone can't see.
A merit matrix is the standard way to do this: performance rating on one axis, salary band position on the other, so a top performer who's low in their band gets a bigger increase than an equally-rated peer already near the top.
Ekaterina puts it plainly: "Performance-by-position-and-range is the gold standard, really,” she says. “The person who is the best performer but lowest in the range deserves a higher percentage increase, and vice versa.”
It's very logical for managers to understand too – we’ve found they actually welcome a framework when you say: these are the rules by which we want to play this cycle."
“The person who is the best performer but lowest in the range deserves a higher percentage increase, and vice versa.”
Ekaterina Potter
Global Director of Reward and Belonging at commercetools
Daniela describes the same priority order: "If there are people who are underpaid and we can see their performance is really good, we definitely want to bring them to the market level or into the ranges – that's priority number one."
One nuance at commercetools is that, whilst they use a structured merit matrix approach, they still factor in manager discretion. “We typically give managers, in each of the boxes in the merit matrix, a range, not necessarily an exact number," Ekaterina says.
"It could be 3 to 4%. It could be 5 to 7%. It could be 0 to 2%, depending on where team members sit in the range and how they perform. We say: this is our salary increase proposition to you and this is the budget cap you have. You may have some context that I don't have, sitting in the total rewards team. You are the manager, you make the final call."
“If they do make any departures from our recommendation we also require a formal justification – not just 'he's a really nice person and I like working with him,' but documentation supported by data and KPIs. That’s critical, especially with the EU Pay Transparency Directive coming into force."
Daniela introduces a mechanism to overcome additional budget requests. "We introduced a wish list," she explains.
"The manager can only spend the budget available to them, but if they can still see there are people who deserve more they can put the person on the wish list with a priority, and a comment justifying the request – and at the top management level, we decide whether they'll get some extra money or not."
As we’ve seen, manager input is an important factor in pay adjustments – but it can also be where the structure and consistency of the cycle falls down.
Here are the two most important problem areas to look out for in Ekaterina and Daniela’s experience:
Both point to the same fix: a defined structure for Rewards recommendations, that’s clearly communicated to managers, and a requirement for managers to document the rationale for any departures. As Daniela puts it: “we need manager autonomy, but within the guardrail of fair treatment.”
With the framework firmly in place, run the performance review process and gather ratings – then calibrate them across managers before they feed into any actual number.
There are different ways of running this, of course.
Commercetools runs performance management as a continuous process rather than a single annual event.
"We have a feedback-based performance management system where you can have a check-in with your manager at any time," Ekaterina says. "Then we have two focal points in the year – the mid-year review, which is more about looking forward, and the end-of-year review, where we look at how the year went.”
Then the calibration process ensures that performance is handled the same across all teams.
“We have a very stringent calibration process and the conversations can get quite uncomfortable as we seek fairness,” Ekaterina explains. “The leadership team of a particular department gets together, and every employee's achievements, strong points, and maybe some gaps are reviewed line by line. It's a very tough exercise, but it helps us stay on top of budget and consistency when we go into the compensation review."
Daniela flags that if your company operates across multiple locations, it’s also important to consider how the factors you use to rate performance translate across them.
“This year, Aeven added a talent potential dimension alongside existing performance to factor in the long-term value of an individual,” she explains.
“That turned out to be very difficult to calibrate globally, especially for Senior Leaders. Leaders from Denmark see high performance as something very different compared to Czech, Filipino, or Hungarian managers – and that's based on culture."
"Leaders from Denmark see “high performance” as something very different compared to Czech, Filipino, or Hungarian managers.”
Daniela Černá
Global Compensation and Benefits Lead at Aeven
Once ratings are calibrated, run them through the mechanism from step 4 to produce actual numbers, check the total against the budget from step 3, and then get leadership sign-off before finalising the adjustments to be made.
This is also the point at which unexpected promotions, if they're in scope, can put real pressure on the total.
“During our last merit cycle we had a 75% spike in demand for promotions, once the merit cycle had already kicked off,” says Ekaterina. “That was a lot of extra cost to absorb.”
"Managers don't care about our finance buckets," Ekaterina says. "There's a people leader under pressure to retain their team, and suddenly they see: okay, the merit cycle is launched – and they treat it as a pool of capital to fix their problem.”
“We managed to maintain our target merit increase and promote those people despite the huge demand, because we harvested savings from attrition and squeezed every euro out of the total envelope, but it wasn't ideal.”
“My main lesson is: don't fight it, manage the behaviour. Learn from each cycle and improve your frameworks to capture what managers will inevitably try to do, even if they don’t factor it into planning themselves."
In 2026 the median salary increase was 5.0%.
That’s any pay rise granted to an employee within their existing job role i.e. from in-role pay adjustments, but not promotions – the median promotion increase is 23%.
There's also, of course, significant variance by function (e.g. the People function had the highest median salary increase last year at 5.3%), and location (e.g. in the Netherlands the average salary increase last year was 6.3%, but in Spain is was 4.0%).
This data is from the 2026 Ravio Compensation Trends report.
Before any manager delivers an outcome to their team, make sure they're properly briefed on more than just the number.
If they've been part of calibration and given real discretion earlier in the process, they'll already understand how the decision was reached – but knowing the framework and being ready to explain it to an upset employee are two different skills.
For Daniela, this begins before the merit cycle even kicks off. "I always run a one-hour session where I present the process and the ways of working for managers.”
“I also use a training tool that enables me to present all the possible scenarios to managers – this is what you're going to get, this is how the tool looks, this is what you'll be asked to fill in. That's very useful when they can see it in reality."
At commercetools, Ekaterina equips managers with more than the rules – they get the actual language to use.
"We provide managers with literal soundbites – bullet points of conversation parts, sentences they can use for different scenarios: the scenario of only getting a lump sum instead of a salary increase, the scenario of not giving someone a salary increase at all," she says.
“This also includes actively coaching against managers reaching for "well, this is what I was told to do" or "this is what the rewards team is asking of us" rather than standing behind the call themselves.”
The conversations that go badly are worth listening to for what they reveal about the process itself, not just the manager's delivery.
Like, at commercetools, managers found it uncomfortable to tell a hired-in top performer they couldn't get a salary increase because they’d already been hired at the top end of their salary band, and pushed back hard on having to have those conversations at all.
As Ekaterina put it: “that’s then a problem that we bring to the next cycle to address in the process.”
"We provide managers with literal soundbites – bullet points of conversation parts, sentences they can use for different scenarios that arise.”
Ekaterina Potter
Global Director of Reward and Belonging at commercetools
Once an outcome has been communicated to an employee, it needs to be made official and processed – not just agreed verbally in a 1:1.
Send written confirmation of the new salary, the effective date, and the reasoning behind it, then update HR and payroll systems (or instruct the relevant teams to do so) so the change actually lands in the next pay run – including any backdating promised.
Once increases are delivered, run a postmortem before the cycle is considered closed.
"We always need to run a retro," Ekaterina says. "We send out a questionnaire to our compensation planners for feedback: what worked well, what do you think we should improve.”
“We always frame it as we'll try our best, but we'll never be completely on top of everything, simply because we don't have limitless resources – but we can always address problem by problem, in different years."
A merit cycle is a compensation process primarily focused on rewarding employees for the previous year's performance. Some companies use "merit cycle" and "compensation review" interchangeably; others use the fuller term because their process also covers market adjustments, pay equity fixes, and sometimes promotions.
There's no single right answer, but it should be a deliberate decision rather than a default. Including promotions in the merit cycle budget means the two compete for the same pool of money, and demand for promotions tends to spike right when that budget is already stretched thin by performance and market increases. Running promotions as a separate process, with its own budget and its own readiness criteria, avoids that collision, but requires building and maintaining a distinct process to replace the informal one a combined cycle currently provides.
Enough to apply context a framework can't see, but within a fixed budget cap and a requirement to justify any change to a recommended figure. Full discretion without accountability tends to produce either budget overspend, or a manager quietly denying an increase to someone who was clearly owed one – both traceable back to a lack of guardrails, not the discretion itself.
That depends on compensation philosophy. Many companies deliberately don't index pay to inflation or the cost of living, instead anchoring increases to local market rates for the role – on the basis that inflation reflects cost of living, not the market value of the work. Whatever the position, it needs to be decided in advance and communicated clearly to managers, since inflation is one of the most common points of pushback once a cycle is underway.
A one-time lump sum instead of a permanent base increase is one option, used to recognise performance without pushing pay above the range. It's not a comfortable conversation for managers to deliver, and it's worth treating as a policy to revisit each year rather than a permanent fix.
No, but a matrix that weighs performance against salary band position produces fairer outcomes than a flat percentage per rating. Without it, people who already sit high in their salary range keep pulling further ahead of equally strong performers who don't, even if both receive the same rating.
Your monthly dose of market insights and expert perspectives
Join Ekaterina Potter (Global Director of Reward and Belonging at commercetools) and Daniela Černá (Global Compensation and Benefits Lead at Aeven) for an honest conversation about running merit cycles in practice – the wins, the parts that broke, and what they'd do differently.
A merit matrix combines performance ratings and salary band position to ensure consistent, equitable pay increases across your organisation – learn how it works, including a worked example.
Vaso Parisinou (CPO at Ravio) and Saime Kutuklu (Director, Central Total Rewards at Delivery Hero) had an honest retro on their latest salary review cycle – the wins, the challenges, and the lessons learned.
