Where I have seen most companies get this wrong is that they do not extend their leveling system into the hiring process, or they do not have compensation tied directly to leveling. Missing either of these leaves huge loopholes in any system.Marco Rogers, veteran engineering manager*
Compensation is a complex subject rife with potential pitfalls. It’s important to develop a compensation philosophy and explore various strategies when building a compensation plan for new hires. Even if you work for a large company with an established compensation plan, you still may gain from reading through these strategies for mapping levels to compensation, as it will help you better understand your company’s and candidates’ perspectives, and will better prepare you to communicate with candidates about compensation.
The best use of money as a motivator is to pay people enough to take the issue of money off the table.Daniel Pink, bestselling author*
When interviewing and assessing candidates, taking a approach helps you map candidates to your existing structure. The goal is to be able to predictably determine their compensation based on the role and level.
Developing an understanding of the talent market in which you’re competing is the first step. For example, early-stage startups may rely more on equity than cash for compensation. Roles that are more senior or more scarce might command higher compensation. Supply and demand really do come into play here. Ultimately, for a particular role—with its associated level of seniority, geographic area, and stage of company—there will be a distribution of what companies offer to their employees (with a high end and low end, and some mix of cash, equity, bonuses, and/or benefits). You can find this type of data through specialized companies that run salary surveys (like Radford, OptionImpact, RHR, or Connery Consulting), on AngelList, through your investors, or even by browsing sites like Levels.fyi, Glassdoor, or Paysa. (The latter group of sources rely on unvalidated, self-reported data, and so may be less accurate.)
The benefit of using data from sources like Radford and Connery is that they’re inherently tied to their leveling ; if your levels correspond to theirs, you’ve got a head start on mapping your compensation to those levels. These data give you a solid indication of what market rates are. The next step is to adjust for your own hiring philosophy and practicalities, and then actually test out your numbers in the market. If you’re losing a big chunk of your offers because of your compensation package, that is the market’s way of telling you something is off.
caution In dynamic markets, large-scale salary survey data can lag as an indicator, and tend to be weighted toward levels in larger companies in larger markets. So you’ll want to factor this in as you consider your own company’s size, maturity, and needs.
Levels.fyi has collected a trove of employee-submitted salary and level data, allowing you to benchmark compensation across levels for a variety of larger tech companies (bearing in mind that these are not vetted in the way that data from formal salary surveys are).
Once you understand the market, you will likely want to consider how or whether your compensation plan might align with your value proposition to candidates, the types of candidates you want to hire, and your company’s constraints. For example, you might consider offering top-of-market compensation to all candidates (sometimes called “the Yankees approach”), in the hopes of increasing your . But keep in mind that offering high compensation won’t hide a bad work environment or lack of vision from new employees.
On the other hand, you might compensate a little more conservatively, hoping to put more of your money into scaling your company. If you can offer candidates a really strong value proposition, like opportunities for growth or a mission they are passionate about, or if they have confidence in how successful your company will be, they might place less weight on cash, and you might offer more in equity.
dangerSome companies can use this strategy exploitatively—but it’s unscrupulous to over-promise on any dimension simply to convince someone to take lower pay than they should. It is likewise short-sighted to pay someone less than they deserve because they are motivated by the company’s mission and might choose to work for less. Others may be motivated by mission but unable to sacrifice making a living.
Ultimately, the best strategy is to compensate as fairly and transparently as possible, and to find candidates whose needs align with what you can offer. If you overpay, you risk attracting the candidates who only respond to extrinsic motivations. If you underpay, you may not be able to hire and retain the people you need to succeed.
Building a healthy compensation system requires associating a compensation band with each .
A compensation band (or pay band) is the range of compensation a company offers to all employees at a certain . Within a band, individual employees’ compensation varies based on factors such as job function, experience, location, and performance.
Again, salary survey data offer good starting benchmarks for this. Companies typically make compensation bands wide enough to have some flexibility, and often these bands overlap. If you want to be really structured, you can peg your compensation to a percentile. For instance, you might decide your compensation will be at some percentile of the data you gathered when developing your compensation philosophy. The 75th percentile is usually a good place to aim for—it’s higher than the average, but not high enough to make your offer “only about the money.”
confusionPeople often confuse percentiles and percents, with serious consequences. When someone recommends compensating at the 75th percentile, they are talking about percentile rank, or paying at a rate that is greater than what 75% of companies pay for a role (in a given market, stage, et cetera). This is very different from paying 75% of what a fair market salary would be.
cautionHow you fit people into roles or levels and how those levels relate to compensation is a minefield for discriminatory practices. Take great care when leveling a candidate and deciding compensation. Under-leveling may result in a weak offer that won’t appeal to the candidate and is also more likely to happen to candidates. Former Google engineer Kelly Ellis and two colleagues filed a class-action lawsuit against Google alleging that they were consistently placed in lower levels than male colleagues, resulting in lower salaries, bonus and stock grants, and promotion opportunities.* Conversely, over-leveling might mean you have expectations for performance beyond what the hire can deliver.*
Veteran engineering manager Marco Rogers wrote a must-read Twitter thread about how he approached mapping levels to compensation at Clover Health, and the pros and cons their approach had when it came to bias, discrimination, and the overall integrity of their hiring practices. (He also discusses salary transparency and non-negotiation policies, which are helpful to understand when preparing to make offers to candidates, and which we will get into next.)