“Things are always happening to me. I’m that sort of bear”
A while back I wrote a bit about the differences between traditional salary structures and global grading systems as a means to manage base pay. In this much-anticipated1 follow-up, I’d like to write a bit about how job-based ranges differ from salary structures, the pros and cons, and the types of companies that use this approach. With pay transparency renewing the focus on salary ranges, it felt like a good time to revisit this topic. I also went on to write about the “hybrid” approach here.
First off, let’s clarify the differences. Having already spoken about the difference between the two most common salary structure approaches, I won’t spend any more time on the distinctions but for the purposes of discussion, I’ll bring back an illustration that indicates what a “Globally-graded” or “Career-leveled” salary structure looks like where a group of jobs e.g. finance, is covered by a single range:
The most common criticism with this approach is the lack of perceived alignment to the market. Common reactions from recruiters and business leaders are that certain roles command higher or lower pay than the structures suggest and that they need job-specific market data to be confident in their pay decisions.
Enter job-based salary ranges:
Under this approach, every unique job within a company’s job architecture has its own range based on market data for the position in question. In my experience, predominantly of US-based tech companies, there is a fairly even split of companies that adopt this approach versus companies that create some version of a salary structure and, as with many decisions, both approaches can work effectively so long as they are used in the right way.
Let’s think about the advantages of job-based ranges first:
- This is by far the simplest approach to explain – It requires almost no upfront explanation. With this model, every job has a range based on market data
- This approach provides, ostensibly at least, the closest alignment to the market. Since the data is the most granular, it should be the most closely aligned to the market and even minute discrepancies in the relative worth of individual roles will be, in theory, reflected in the range
So on the face of it, every company should use job-based ranges, right? Well, in this author’s opinion, not quite. To those of us working with job matching and market data for some time, some flaws in this approach become apparent:
- Market data is far from perfect. Compensation data is built from a database of multiple companies and employees with new companies joining, old companies leaving, promotions, RIFs, new hires, salary adjustments all happening, all the time. As such, the more granular we get, the more of this “noise” we are exposed to to the point where it can be impossible to differentiate the noise from the trend. In my experience, about a third of the jobs in a given survey will move down and not up in a typical year of economic growth. Often the year-over-year changes in roles will vary between -5% and +10%. Does that mean we cut pay for those downward-moving roles and give others a 10% increase? These changes are often not reflective of a trend and simply noise in the data, but translating that data into practical updates requires work
- In addition to market data being imperfect, it isn’t always available. In the US, and a handful of other western countries, job-specific market data is readily available but even in these locations, there are frequently jobs that do not have enough market data behind them to populate and we have to find creative ways to plug the gaps. Outside of this handful of countries, job-based benchmarking is simply not feasible meaning that other salary structure approaches must be taken
- Using job-based ranges can create a huge, possibly overwhelming catalogue of salary midpoints in larger companies making the process of updating ranges more complex. Given the noise in the market data, making regular updates to the ranges is not as straightforward as one might first assume. This is usually a fairly involved process of reviewing each range against the updated market data, and building in parameters to allow for market movements without allowing moving midpoints too aggressively
- There needs to be a high degree of confidence in the underlying job architecture, and job matching. Choosing the appropriate range, is contingent on choosing the appropriate job which, as many of us already know, isn’t always simple. With no functional or broader ranges to fall back on, every individual needs to be aligned to a specific job which means more time spent understanding roles, discussing with hiring managers and recruiters, and updating internal job architecture to ensure the appropriate fit. In addition, once hiring managers “figure out” the high-value roles, you might see a trend towards matching to higher value roles in order to unlock higher pay. So although simpler to explain at the outset, there may be more insidious challenges down the road
- New roles require new ranges. If a role is being created that doesn’t have a range, someone needs to create one. While this may be less of a challenge in a large established company, in a fast-paced recruiting environment, this can be challenging, particularly if the role doesn’t have data available and the HR team is small
- We shouldn’t overestimate our ability to manage pay. In every company, there are outliers, exceptions, and one-off “special” cases that don’t fit into salary structures. Yes, job-based ranges have the potential for closer market-alignment but it is important to be realistic about how effectively we can manage pay to those ranges and the degree to which individual skill-sets and backgrounds create variances
The final point I will make on this is that quite often, when noise is accounted for, the differences in value between roles is simply not as severe as many would expect. Salary structures are not intended to account for short-term lumps and bumps in demand for certain skillsets. Built in the right way, a salary structure can easily accommodate roles of differing values without the need for job-based ranges and should allow for short-term nuances in the availability and price of jobs in the market. Training hiring managers and recruiters on the nature, challenges, and trade-offs of job-based market data can help dispel myths about the validity of a salary structure approach.
So when is the right time to use a salary structure vs. job-based ranges?
The most obvious contenders for job-based ranges are more mature companies in industries that have access to reliable job-based market data. An easy example is big tech in the US where many large companies use this approach effectively. There are a couple of reasons for this. The first is that the job architecture catalog is widely understood and relatively static. When new individuals are hired, the chances are they will fit into the existing architecture without the need to develop a new job or range. The second reason is simply time and resources. Larger companies have teams of experienced comp professionals, and experienced HRBPs to help navigate job matching conversations, salary range updates, train hiring managers, and benchmark roles.
By contrast, a small high-growth company with limited resources is likely to add new jobs on a regular basis meaning that job-based structures will require regular updates, job matching, and leveling conversations to understand the nature of new roles with a team of business leaders that likely don’t have much time for these conversations. As such, I usually recommend a level-based approach for small high-growth firms.
As with so many things, there is no right or wrong approach, so long as the limitations and challenges are properly understood. The chosen approach will be a function of familiarity, business sentiment, growth plans, global coverage, underlying architecture. Since pay is a hygiene factor, my personal preference is always to lean towards simpler approaches which frees up HR’s time and resources to focus on initiatives that will truly drive greater engagement.
1Not anticipated at all. By anyone.