Global Pay Strategies: What to do When You Don’t Have Salary Data

The last few years have seen a transformation of workplaces that rivals any of the industrial revolutions. Catalysed by the pandemic, workplaces took up a level of global distribution unlike anything seen before. A confluence of the hyper-competitive talent market meeting the forced, but rapid uptake of distributed work, re-wrote corporate sentiment to where and how people could deliver their roles. Companies quickly ascertained that, for knowledge workers, borders didn’t constrain exceptional talent, and the best companies would win by having the best talent. In 2021, you’d be right for thinking we’d all be remote forever, and as a result, many organisations doubled down on distributed work by looking abroad for skills. While we’ve since seen a partial snap back to colocated work, it’s still safe to say distributed workplaces are here to stay. But with this phenomenon comes new challenges. With talent being borderless, companies are employing their people from countries further and further afield, and with an increasing lack of salary data to inform their efforts. So for these companies, the question prevails: what do I pay people in the countries with little to no salary data?

While the number of compensation data providers only continues to blossom, many of these vendors remain focused on deepening their dataset in the narrow, densely populated markets that their customers primarily employ. This means for employers and workforces in predominantly Westernised and developed nations, the challenge is reduced by the bevy of vendors to choose from — great! But if you’re an organisation employing in more far-flung places, there are some alternative methods you’ll need to explore for adopting an informed pay position. In the past, I worked for a distributed organisation that could (and did) employ anywhere in the world. During my tenure, we scaled from sub-200 to over 650 people, in less than 12 months. Without location constraints, we didn’t know which country our new hire would be based in until we went to make an offer. Quite the compensation quandary, but one that led to some deep experience in handling this challenge. With that said, let’s explore three considerations, ranging in flexibility and maturity. Which approach is right for you will depend on several factors, and we’ll explore those alongside each method.

The ‘pay them what they ask’ approach

When a startup, ahem, starts up, they don’t often have the robustness of thought around salary that a more mature and established organisation does. Often, the founding team will have a loose idea of what they want to hire, then rely on referrals, their VC’s, or pull together job advertising in a bid to rapidly build capability and get their product to market (or growing) as quickly as possible. This often results in an approach where pay is based on what the candidate asks, and the recruiter or Founder just uses a gut check before offering. This speed to market makes total sense in the moment because they have the cash to burn and the capability to build. And it’s understandable, the path of least resistance then is to pay the asking price and deal with the downstream consequences, or people-debt, later. And boy do they. But before we get there; what are the benefits of this approach when we think about it in a global setting? Especially when we have little to no way of ever knowing whether the amount a candidate is asking for or the company is paying, is right in the context of our organisation's pay strategy.

The pros:

• Attracting top talent - If you’re happy to meet the expectations of candidates no matter where they are, you have the unique opportunity to hire high-quality talent from around the world. This can deliver access to a diverse pool of skillsets and expertise for your company.
• High acceptance rate - If candidates are getting what they ask for, they’ve got no reason to decline. However, depending on what the negotiation culture is like, they may be left feeling like they left money on the table.
• Improved candidate experience - Everyone loves getting what they ask for, so they’d no doubt enjoy receiving a job offer with a salary to match their expectations, and because they’re satisfied that they’re being offered what they’re worth.
• Hiring agility - If you’re taking this approach then you’re only rejecting candidates based purely on performative reasons, and not for their misalignment to budget—this means more candidates to evaluate. Additionally, you’re not dealing with an often lengthy negotiation stage, meaning you’ve got the broadest talent pool and a faster hiring process.

The cons:

• Prone to overpay - At the end of the day, pay data exists for a reason, and it’s to help companies control their finances by informing them of the market rate for skills. Overpaying in any market is a cash-burn killer for startups with a finite runway.
• Prone to underpay, too - If you don’t know what the right salary is, the candidate probably doesn’t either—maybe you’re perpetuating salary suppression, or the candidate realises shortly after that they undervalued themselves and head off for a higher paying job.
• No yardstick for negotiations - employees may ask for increases and it's entirely unclear what is within reason, so you feel obliged to pay it with no defensible way to push back, lest you risk attrition.
• Pay equity - Eventually (maybe by the end of this article!) you will likely establish a more robust method for gauging the value of talent in disparate regions. The longer you employ this method the more you hurt your chances of ever seeing pay equity be achieved.

Personal note: while this approach is a necessary part of starting a business, I would urge you to move past it as quickly as possible. In my experience managing precisely this scenario, the longer you are in a stage where there are no checks and balances, the greater the disparity and debt you create when you do introduce structure and find that you have populations of people overpaid and underpaid. Worse, those massive pay inequities tend to come due during a crucial growth stage and when money is sparse.

The geographically agnostic approach

While not a new concept, with the increased prevalence of distributed work came the argument of equal pay for equal work. A term once reserved for gender disparity, it is now applied to the geographic trait of employees. The argument is; why should someone doing a job in a typically low-cost region be paid less than someone doing the same work in a high-cost region? That’s a very philosophical debate that I won’t get into for this piece. One way in which you can leverage this approach is by using salary data you do have access to, to open up the talent pool. Say you are a Canadian start-up and you know all aboot the salaries in your region (I had to do it). You’re finding it harder to find the skills you need locally, and so you take that salary information and use it to hire abroad. What this does is tie you to a point on the global salary spectrum, and allow you to hire from locations where the salary is typically at or below the amount you’re offering. You may ask, ok, but how do I know which places pay more or less than Canada? Well, in some ways you don’t have to—you can use the prevalence of remote job boards, be upfront with your salary, and those who are prepared to accept that amount will apply. But if you’re looking to take a more strategic approach, a good rule of thumb can be the Cost of Living index. Now it’s not perfect, but using resources like Numbeo, you can determine approximately where on this spectrum you’d land. Arguably, countries below the benchmark country would be ones in which you will have an easier time hiring because the salary would be more competitive than the typical market rate. The example below gives you an idea of where it would be hard to hire (above Canada where it's more expensive) and easier to hire (below Canada where it's cheaper).

Now, you may find that remote workers in regional Austria find that salary very attractive, and that would be an example of you still being able to hire in typically higher cost-of-living locales. But what other implications are there?

The pros:

• Easily scalable. Using the data you do have available, you now have a scalable salary approach to hiring talent. You don’t have to buy additional salary data, and you don’t have to do salary research anywhere but in your benchmark/target country.
• Greatest focus on pay equity. Any focus on pay equity can be maintained without interruption. You’ll foster inclusion and be a highly desirable workplace to those who value fairness.
• Easy global mobility. If your people move anywhere in the world where that salary is competitive, you’ve solved a very challenging global mobility problem.
• Feels good. This is a macro effect, but you’re creating the opportunity for talent to exist outside metropolitan/geographical centres—which can be a nice touch.

The cons:

• Limits access to the high-cost side of the talent pool. Depending on your benchmark country and where it lands on the global salary spectrum, you may be restricted from accessing talent in more expensive locations. This may be exacerbated with the more niche or specialised skillset you look for, as they might predominantly reside in costly locations due to the high salaries that can be paid there. If you were to target those higher-cost locations, you’d be increasing your salary bill for all talent due to the agnostic approach.
• May lose people moving to high-cost locale. If one of your people moves to a high-cost locale, you might find the salary is no longer competitive and you’re unable to retain them.
• An equitable base is easy but equitable total rewards is hard. While you may be equitable on base compensation, statutory requirements may mean you are inequitable on total compensation. For example, if you hire someone in Australia, where there is a mandatory superannuation (pension) contribution of 11% of base salary, what do you do for those in Canada where you don’t have a similar obligation?
• Overpaying in lower-cost regions. Technically, you’re paying more than you need to in any location below yours on the spectrum. If you’re a business that is growing rapidly or has lower margins, it can be challenging to ignore that as a cash optimisation.
• May lose existing talent in high cost regions. If you already have people in a location that was more expensive before adopting this model, their salaries may have to freeze and you may risk attrition.

While this isn’t necessarily a con so much as a watch out—you should be aware of the minimum wage laws of the countries you’re hiring in. Of all the salary data you might have trouble finding, minimum wages should be the easiest.

The geographically localised approach

On review, you may find yourself philosophically or commercially opposed to taking the geo-agnostic approach outlined above—why? Maybe you want to access talent in the world's most expensive regions, but you can’t afford to offer that same pay to every other country. Maybe your profit margins or financial conditions prohibit you from taking this approach. You may also be uncomfortable with the extremes that a geographically agnostic approach would represent, with the potential for a high salary at one end of the spectrum and a low cost of living at the other. Let me give you an example. Let’s revisit Canada; if you take the average salary (50th percentile - CAD 65,000), and apply it to a traditionally much lower-cost country like Indonesia, it converts to ~750 million Indonesian Rupiah (IDR). Fine, but what is a normal salary for someone like this in Indonesia? Well, we’re probably hiring a skilled knowledge worker, so let's aim for a 90th percentile salary—here we get approximately 100 million IDR. This means our Canadian salary represents 7.5x what is already a top-tenth percentile salary in Indonesia. The first question that tends to come up is whether that is responsible, and what kind of implications it creates for that individual from a social standpoint. Regardless, here we’re exploring a more nuanced approach to the salary data equation. There are a couple of different methods to how we can approach this: one is bundling, almost like a micro-geo-agnostic approach; and the other is using a more granular method.

First up is the bundling method

Imagine you’ve got people employed across ten different countries, some in high-salary locations like the US and Canada, and some in lower-cost locations like the Philippines and India—two distinct salary cost groups. You may find that your salaries for these groups are similar internally, and you may be able to access salary data for one or two countries in that bundle. You can then use the knowledge you have of these markets to effectively group those countries and apply a consistent set of salary bands to that group. This would mean for a single role at a single level, you might have two salary bands. One for the higher salary cost geographies and one for the lower. This method enables you to more granularly apply a set of principles where the salary might not differ much locally, or where it's operationally more efficient (and arguably cost-negligible) to do so.

The second is the granular method

This approach is more granular, and uses a recognised and widely/freely reported measure, such as COL, to dictate where on a salary spectrum a country might sit. Here you can take the countries you have, place them in a spectrum from highest COL to lowest COL (using something like Numbeo), and determine the range and approximate distance for each country between the top and bottom. Using the salaries of your current employees and any data you have for those markets, you’re then able to start generating a salary point at each of the intervals those countries sit using a method such as linear regression. Eventually, you could end up with something like this.

For illustrative purposes only

It’s important to mention here that COL gets a bad wrap, and I understand why. It’s something that is reported broadly and yet is felt individually. My COL as an individual could be wildly different from my neighbours’, and yet what's reported is the country average. The price of a ‘market basket’ of goods is also hugely subjective. For example, a 2-litre bottle of Coke in Lebanon is the equivalent of AUD 13, yet in Australia, I could buy it for $3—the COL in Australia is higher than in Lebanon though. COL is also not a perfectly linear representation of salaries, which are driven by market forces. But, when we’re trying to come up with a robust method for establishing a fair salary to be equitable in our approach, it’s a start. But it does need your attention over the long term to ensure it is aligning with your compensation philosophy.

The pros:

• Pay is highly relevant to the location. You have an opportunity to control your compensation approach to more granular geographic levels and use a data-driven approach to inform scalable hiring across the globe.
• Strong guidance while maintaining flexibility. Deeply flexible and customisable to your organisation's needs. Need to be more competitive in a region with a higher concentration of desirable skills? Move it up the spectrum.
• Most cost-effective for your growing business. This is your best approach from a cost optimisation standpoint. By being able to offer salaries that are competitive locally rather than tied to other regions, you can reduce spending without compromising access to talent.
• Globally relevant. If permitted, those who do move globally can rest easy knowing their salary is being re-determined for their new location, competitively.
• Consistency equals happy employees. By taking a consistent approach, employees can be confident and comfortable that their salary is determined in line with others across the globe.

The cons:

• Heavy lift. At its most granular, it will require the most work to keep up to date, and that comes at a cost of time and effort.
• Still imperfect. You need to fact-check these salaries in the market, for example, what are salary expectations in that region? What are your current employees in that region being paid? What is your offer acceptance rate? etc. These are all ways you can inform whether the model works or needs to be adjusted.
• Requires good data governance. Because the linear regression method is a mathsy approach that needs mathsy considerations; if you have outliers in your dataset (for example you’ve offered abnormally high or low salaries to your employees) then your model may be off. Consider if they need to be removed (see the point above about adjustments).
• Doesn’t solve for digital nomads. If you are a proponent of digital nomadism and permit your employees to travel the globe, those in a higher COL region are going to be advantaged because they have a higher salary with which they can travel the world. Those in a lower COL region will have the opposite.
• May incentivise levelling up geographically. You may need to put controls in place as to how you support global mobility for your workforce, as your people may now be incentivised to relocate internationally for a higher salary.

As you can see, no approach is perfect, and each has its benefits and drawbacks. As companies either newly or increasingly adopt global remote work practices, the need for an informed and measured approach toward what to pay is critical. While companies can initially be fast and loose with a ‘pay what they ask’ approach (assuming their financial situation permits), they should move quickly through this phase if they are to mitigate the longer-term impacts and people debt they will inevitably incur. This can be done by adopting either a geographically agnostic or locally relevant approach to their compensation. Each should be assessed on its own merits in tandem with the unique needs and strategies of the business.