ReportSee where pay is heading in 2027View report

How to choose the right peer group for salary benchmarking

Benchmarking

You’ve rebuilt your salary bands using the latest salary benchmarks, you’re presenting them to the leadership team, it’s going well.

Then someone asks why the comparison doesn’t include the biggest names in tech, or the market leader in your sector.

It’s a familiar moment for Juan Novoa, Consulting Director and Founder at Valor Consulting.

“Stakeholder expectations on what peer group you should be comparing to often derail the conversation,” he says. “Sometimes leadership wants to see the company as competing in a certain market, when that isn’t the reality of where your talent comes from and leaves to.”

Benchmark against the wrong peer group and your bands can miss in either direction: more expensive than the talent you need, or not competitive enough to win it. 

The data still looks precise – it’s just precise about the wrong market.

We spoke to Juan and Andreea Lupu, Founder and Fractional Compensation and Benefits Architect at Meaningful Consulting about how they go about choosing the right peer group for a company, how big it needs to be, and how they get leadership to agree with it.

Why is defining your peer group important for compensation benchmarking?

Benchmarking exists to answer one question: what do you need to pay to win the talent you need, without paying more than you have to?

Your peer group is what makes sure the benchmarks you use accurately reflect “the talent you need”. 

Benchmarks only reflect your market if they’re drawn from the companies you actually compete with, the ones you hire people from and lose people to. 

Compare against the market as a whole, or the wrong peer group, and you’re pricing a different talent pool.

Venn diagram titled “Who belongs in your compensation peer group?” showing the overlap between companies leadership want to compare against and companies you actually hire talent from and lose talent to.

In practice, that means defining your peer group by the factors that shape pay: location, industry, company size, and funding stage.

“I’ve worked with clients who have the philosophy of saying, yes, we want to be upper quartile – but upper quartile of what market?” says Juan. “Depending on that market, you could be upper quartile, upper decile, or you could actually end up paying the median.”

The difference between benchmarks using different peer groups can be large. 

Take funding stage as an example. 

For a mid-level (P3) Product Manager in the UK, Ravio’s benchmarks show the median base salary is £66,800 among early-stage companies and £72,700 among late-stage companies – a £5,900 or 9% gap for the same role, level, and country.

In reality, you’d usually combine several factors, such as early-stage fintechs in the UK, to get closer to the market you actually compete in.

Bar chart of median base salary for a mid-level UK Product Manager by funding stage using Ravio's benchmarks: £66.8k early stage, £69.7k growth stage (+4.3%), and £72.7k late stage (+8.8%).

How to define the right peer group for your company

Defining your peer group comes down to four steps, starting with your compensation philosophy and ending with how you build it in your benchmarking tool. 

Here’s a quick summary before we go through each one:

Four-step guide to defining your peer group: start with your compensation philosophy, gather data on where you hire from and lose talent to, decide if and when to use a different peer group, and build it with your benchmarking provider.

1. Start with your compensation philosophy

Who you compete with for talent is a strategic decision, never just a setting in your benchmarking tool.

It belongs in your compensation philosophy, alongside where you want to position pay – which means that’s the very first step in defining the right peer group.

“It’s important that data doesn’t dictate the philosophy, but the other way around,” says Juan.

Four questions give you a starting point:

  1. What market are we competing in?
  2. Where do we want to sit in that market?
  3. What’s the balance between base and variable pay?
  4. Are there roles where we’d flex that position?

Because it’s a strategic decision, leadership should help make it, rather than just hearing the result when you present your bands and questioning it then. Once leadership has agreed who you compete with, the benchmarks follow from a decision they already bought into.

If views differ, bring it back to risk.

“When you talk about the risks – of paying too much, of not being efficient, of having to answer uncomfortable questions from board members – that conversation tends to change,” says Juan.

And if you don’t have a philosophy yet, start with an imperfect one, and keep iterating and improving on it over time.

“A compensation philosophy isn’t a standalone event or project,” Andreea explains. “It’s something you’ll revise probably every year.”

2. Gather data on where you hire talent from and lose it to

Defining your peer group first means looking internally at who your talent competitors are. 

“You should have a very close relationship with your recruitment team and hiring managers, because they have information directly from the candidates who applied to your company, and who were lost at the offer stage,” says Andreea.

Exit interviews show the other side: where your people go when they leave.

“Normally you find you’re looking at your closest business competitors – companies with similar operating models, funding structures, and recruiting pools,” Juan says.

3. Decide if (and when) to use a different peer group

You may find that there are multiple peer groups for your company. 

Different job families can compete in different markets, for instance, and if your company hires across multiple locations then those geographies may also contain different talent competitors. 

Juan and Andreea recommend that there should be one core peer group across the company, which is what your salary bands and pay structures align to. Keeping this consistent helps to make pay decisions explainable. 

However, some scenarios or requests will always call for a different comparison.

“I’ll first ask: what is it for? What triggered the need to recheck the benchmarks or peer group?” she says.

Juan gives an example of a one-off that justifies a narrower sample.

“If you’re a retail company and your legal team challenges the pay for a senior legal advisor, because you often recruit legal advisors from mid-sized law firms rather than other retail companies, there could be a valid case for using a benchmark from law firms to inform the conversation,” he says. 

“But you’re not going to benchmark every role across the business to law firms just because of that.”

So it’s fine to reflect those multiple peer groups in your benchmarking and pay structures, as long as it’s a deliberate choice with a good reason, and that choice is clearly defined and documented – ensuring every exception is defensible in future.

“Be very clear about how you benchmark if you use different samples or different approaches for different purposes,” says Juan.

4. Build that peer group with your benchmarking provider: named companies or market filters   

How you actually build a peer group into your benchmarking process depends on the compensation data source(s) you rely on.

Traditional salary surveys typically require you to purchase specific survey cuts per industry and location to align with your peer group – or to pay extra for a custom peer group report built from specific named participants. 

One thing to be aware of with custom peer group surveys is that they can only cover the companies that agree to take part, and the lengthy survey collection and release process means that younger, fast-growing ones will often be left off the list.

Modern compensation benchmarking platforms typically give you market filters in-platform to play around with, so you can filter the benchmarks by factors such as location, industry, headcount, and funding stage, and see how much each filter changes the number.

In Ravio, you choose a position and location, then narrow the market by industry, funding stage, and company size. Each filter updates the benchmark, so you can see how much it moves the number before you commit to it.

For some sectors you can skip the filters and start from a curated peer group instead. 

Fintech Select and AI Select are named sets of companies that Ravio maintains, so the benchmark stays current as the group changes.

Ravio’s market filters and peer groups: location, industry, funding stage, and headcount, plus the curated peer groups AI Select and Fintech Select.

If you’re using filters in a platform like Ravio, Andreea suggests working from broad to narrow.

“First look at the country, then the company size, then the industry,” she says. “And with industry, ensure you’re clear on what the role actually means in your specific industry – a senior engineer at one software company isn’t the same as a senior engineer in another industry.”

How big should your peer group be?

Every factor you add into your peer group brings the comparison closer to your company’s reality. But, it also removes companies from the sample, which might result in less robust benchmarks.

“It can be tempting to use every single filter to get the exact comparator,” Juan warns. “But that can leave you with much smaller, less reliable samples, and some of those filters don’t have a big impact on the market benchmark anyway.”

“Normally, I look for a minimum of 10 to 15 companies in a peer group,” says Juan. “That tends to control most of the variability, especially in niche industries.”

“With fewer than that, you might have to roll up some of your criteria – by region, by company size, or by funding stage.”

“If I can’t find that role in Romania, I’ll look at a similar country, like Italy,” Andreea says. “I can compare salaries in Italy with Romania, and I know they’re around 10% higher or lower.”

She does the same with other factors like industry, company size, or job level – broadening the benchmarks, and then applying a differential informed by roles where there’s a larger sample size.

Say you want to benchmark a mid-level P3 Product Manager against the UK’s top fintechs, but that cut is too thin to rely on.

You could start from the broader UK tech benchmark, which puts the role at £69,700 (Ravio benchmarks, November 2026).

Then apply the premium where there’s more data. Ravio’s analysis of the world’s top 100 fintechs found they pay Professional roles in the UK 22% more than the overall tech market.

That gives you an estimate of around £85,000 – a starting point to sense-check, not a benchmark in its own right.

Where peer groups leave the sample thin, Juan also recommends cross-checking against multiple data sources for confidence – which may mean being upfront with your stakeholders and asking for additional budget.

“We’re allowed to go to leadership and say: this is the data I found, I don’t quite trust it yet, so we need to work together and see what the company can afford,” says Andreea.

Diagram showing how each added factor narrows a benchmarking peer group, from UK to 100-500 employees to Fintech, with a reminder to aim for at least 10-15 companies in the sample.

How often should you review your peer group for salary benchmarking?

Once it’s agreed, your core peer group shouldn’t change often.

“Peer groups tend to be relatively stable, unless there’s an acquisition or a pivot in your organisation,” says Juan.

A change in funding stage, or a move into a new country, is also a reason to revisit it.

Between those moments, watch for signs that your bands have drifted away from the market. Andreea looks for two.

The first is in the bands themselves.

“I look at the extremes in a band – people at the beginning and at the end – and try to understand why they’re there,” she says.

Sometimes there’s a simple explanation, like a recent hire or someone about to be promoted. But it’s worth digging in when people keep clustering at the top or bottom of a band with no clear reason.

The second comes from hiring. When a role is hard to fill, managers sometimes agree to a salary above the band without going through Rewards.

“It needs to be flagged,” Andreea says. “It might be a certain team’s decision, but you’ll see it trigger other events company-wide that you may not like.”

Either sign can point to one of two problems: your benchmarks are out of date, or your peer group no longer matches the market you’re hiring in – so refresh the benchmark behind your bands, and then revisit the peer group if needed.

In summary: build your peer group around who you compete with for talent

Define your peer group in your compensation philosophy, build it from evidence of where your talent comes from and goes to, and agree it with leadership before anyone sees a number.

“Be clear on the principles and the philosophy,” Juan says. “If you’re clear on those, every decision that follows is a lot easier, and you’ll have much better conversations once everyone has agreed on them.”

Keep one core peer group, and document any exceptions. Work from broad to narrow, and always stop before the sample gets too thin to trust.

Building a peer group for your next benchmarking cycle?

Explore Ravio’s benchmarks

FAQs

What is a peer group in salary benchmarking?

A peer group is the set of companies whose pay data you compare your own salaries against. It should reflect the companies you actually compete with for talent: the ones you hire people from and lose people to. The choice changes the benchmarks you use. For instance, according to Ravio’s data, UK fintechs pay a mid-level Product Manager a median of £72,100, against £69,700 across all companies.

How do you choose a peer group for compensation benchmarking?

Start with your compensation philosophy, and agree with leadership which market you’re competing in before you look at any data. Then use evidence from recruiters, hiring managers, and exit interviews to identify where your talent comes from and goes to. Build the peer group from broad to narrow – location first, then company size and industry – and keep only the filters that meaningfully change the benchmark.

What criteria should you use to define a compensation peer group?

The most common criteria are location, industry, company size, and funding stage, because each one shapes what companies pay. Location sets the labour market, company size reflects the scope of a role, funding stage reflects what a company can afford, and industry captures any sector premium. Ravio, a real-time compensation benchmarking platform, lets you filter benchmarks by these criteria or use curated peer groups for sectors such as fintech and AI.

How many companies should be in a peer group?

Aim for at least 10-15 companies, according to Juan Novoa, Consulting Director and Founder at Valor Consulting. That tends to control variability, especially in niche industries. If your sample falls below that, roll up your criteria by region, company size, or funding stage. The right number also depends on how much pay varies for the role, so check your provider’s confidence indicators too.

Should you use the same peer group for every role?

Use one core peer group across the company as your default, because consistency makes pay decisions easier to explain. Some job families or locations compete in a different talent market and may need their own comparison, such as engineers who regularly get offers from much larger companies. That’s fine, as long as it’s a deliberate choice that you define and document.

Should you benchmark against big tech companies?

Only for roles where you genuinely compete with them for talent. If your hiring and attrition data shows people joining from or leaving for big tech, include those companies in that role’s comparison. Otherwise, benchmarking against much larger or later-stage companies will only inflate your bands and payroll costs.

How often should you review your peer group?

Review your core peer group after an acquisition, a pivot, a change in funding stage, or expansion into a new country. Otherwise, it should stay fairly stable. In between, watch for warning signs, such as employees clustering at the top or bottom of a band, or managers regularly hiring above the band. Either can mean your benchmarks are out of date, or your peer group no longer matches your market.

How are executive compensation peer groups different?

Executive peer groups tend to be more formal, and at listed companies they’re often broader. That’s because shareholder return measures bring in companies that investors compare you to, not just your competitors for talent. Read more in Ravio’s guides to executive compensation at startups and executive compensation at public companies.

What is a custom peer group survey?

A custom peer group survey is a salary benchmarking report built from a specific list of companies you choose to compare against, rather than a general industry or location cut. Traditional survey providers typically offer them for an extra fee, and they can only include companies that take part in the survey, so younger, fast-growing companies are often left out. Real-time benchmarking platforms like Ravio offer an alternative: market filters you can combine to build a peer group, and curated peer groups for sectors such as fintech and AI.

Get the Compensation Review straight to your inbox

Your monthly dose of market insights and expert perspectives

You might also like