How to scale a compensation program
Dec 08, 2020, 10 min read
Compensation is an uncomfortable topic. Friends don’t ask how much you make at dinner parties. Parents often don’t tell their children about their income, spouses sometimes don’t tell their partners. You don’t see any paystubs posted on Instagram. And co-workers rarely share their own pay information with colleagues at the office.
This ‘money taboo’ we experience in most cultures relegates the field of compensation to a dark science, filled with mystery, myth and lore. It’s not broadly discussed or debated, and often compensation decision power resides in the hands of few. When this power is harnessed well, compensation can be hugely impactful, driving optimism, productivity, motivation and retention. But when handled poorly, it can be, at best, a waste of money and, at worst, riddled with unfairness and bias.
The ‘money taboo’ we experience in most cultures relegates the field of compensation to a dark science, filled with mystery, myth and lore.
I’ve had the chance to see several companies and compensation programs evolve. Between Salesforce, Uber, advisory work with more than a dozen startups, and now Brex, I recognized a few common patterns that led to success and failure. I hope these learnings can help new startup founders navigate the challenges and timing in building their own programs.
When Compensation Goes Wrong
Let’s first talk about when compensation doesn’t work. And how startups sometimes end up on the wrong path or in the wrong place. And how, unfortunately, this happens more often than not.
When you’re a new startup CEO, compensation is easy. Why? Because you don’t have much money and you don’t have many candidates. So the problem is naturally solved by doing whatever you can to close those first few team members. This is typically a combination of laughably low cash, a meaningful percentage of the company’s equity, and the legally-allowable assurance of an any-day-now multi-billion dollar valuation.
This compensation strategy is sometimes referred to as ‘market of one’. Meaning it doesn’t matter what your intended role is or your past experience, we’re going to pay you whatever it takes to close you just above your reserve price. The next hire, who might be doing a very similar job, could get a totally different compensation package because we did whatever it took to close them just above their reserve price. And so it continues for the remainder of the founding team. Because everyone has different reserve prices, compensation is extremely inconsistent for these early employees.
Some startups put their heart and soul into it, but unfortunately don’t make it. The company, and all their compensation messiness, dissolves inconsequentially into history. But other startups do make it. Their business model shows promise, a fresh funding round is rewarded, and the growth continues. This is an important moment. This is the critical crossroads for the long-term stability and success of a compensation program. These next hires should no longer be considered and compensated as part of the ‘founding team’. Instead, new hires should now be treated as part of the long-term company, and a sustainable compensation philosophy should now crystalize and apply to these new hires.
Compensation can be hugely impactful, driving optimism, productivity, motivation and retention.
Startups that don’t make the transition and continue to hire with reckless abandon for compensation may one day find themselves in a precarious situation. Without structured compensation, each new hire, promotion, and transfer will exacerbate the inconsistencies. So when does this become a problem? Over time, the differences in compensation treatment will become more and more apparent. Systemic gender and ethnic biases will manifest. Managers will see the inconsistencies within their teams, rumors will begin to spread about who makes what, employees will begin to ask questions, management won’t have answers, distrust in management will form, and people will ultimately depart to seek fair compensation and management elsewhere.
And like technical debt, each day that passes makes it increasingly harder to correct the underlying issues. Fixing compensation often means establishing the target pay for each role and level and then moving individuals up or down to get them to the right place.
And like technical debt, each day that passes makes it increasingly harder to correct the underlying issues. Fixing compensation often means establishing the target pay for each role and level and then moving individuals up or down to get them to the right place. This is hard enough for 50 people. But imagine the effort, emotion, and expense that comes with fixing it for 500 people or 5,000 people.
I once had to reset a comp program for a company with 15,000 employees. After almost a year and nine figures of corrective dollars later, I can confidently say that establishing a comprehensive compensation philosophy earlier is better than later.
Building a Scalable Compensation Program
At Brex, we did start early. After our first significant funding round at ~50 employees, we began creating our compensation foundations. This included an early start at ‘leveling’ where we connected job responsibilities to a specific compensation band for each role in each organization.
To ensure that our compensation bands were correct, we continuously revised them based on qualified competing offers presented to our recruiters. For example, if a software engineer had a competing offer at 5% more than our current target compensation, then we matched the offer, hired the engineer, and then moved all other engineers at the company to that new benchmark compensation number.
If a software engineer had a competing offer at 5% more than our current target compensation, then we matched the offer, hire the engineer, and then move all other engineers at the company to that new benchmark compensation number.
This practice helped us stay competitive while simultaneously tuning our compensation ranges to the market position best aligned with our company strategy. More recently, we standardized this approach by using compensation benchmark surveys provided by external partners (we use Radford). This data is provided to us twice per year and we use it to drive market compensation adjustments in each of our compensation cycles.
Another early compensation principle we built at Brex focused on providing as much compensation flexibility as possible for our team members. Early employees sometimes asked to adjust their offer mix — sometimes reducing equity in exchange for higher cash, sometimes vice versa. We quickly recognized that everyone is in a different financial stage of their life and that everyone has varying levels of risk tolerance.
We began offering employees a choice. For a role where the total compensation (inclusive of cash, equity, and bonuses) is, for example, $100K, a new hire can choose, with some guardrails, whatever portion they’d like in annual cash and whatever portion they’d like in annual equity. It’s a scary thing to attempt because Finance and the Board need to be able to accurately forecast cash burn and equity dilution. But, as it turns out, it’s a fairly predictable thing.
Brex employees, on average, choose about 32% of their compensation in equity. As expected, equity election consistently skews higher with seniority and experience. Those in the top quartile of compensation choose nearly half of their pay in equity. But there are still high deviations within levels and roles, which tells us that a high-flexibility approach optimizes for the individual financial circumstances of our employees. This unique strategy has worked for us. We’ve seen no meaningful correlation between equity election and performance markers such as attrition or promotion. And, as an added benefit, we’re now able to attract talent to Brex who ordinarily would avoid startups because of liquidity needs.
We’re now able to attract talent to Brex who ordinarily would avoid startups because of liquidity needs.
The Value of Predictable Compensation
This cash/equity election experiment inadvertently created another compensation principle that we realized our employees valued: Predictability.
An offer from most mid-stage startups includes a complicated menu of pay elements such as Base Salary, Bonus Target, Performance Multiplier, Company Multiplier, New Hire Grant, Annual Equity Refresh, and Merit Increases. And each one of these components includes rules and criteria that, when combined, make it nearly impossible to predict your actual annual compensation.
The traditional assumption is that pay connected to performance incents individuals to increase their productivity, thereby unlocking incremental individual and company bonus value. This creates the need for complex performance management systems that discretely measure an individual’s contribution relative to others and price an associated reward.
Administrative cost of performance measurement exceeds the benefit of incremental employee productivity.
At Brex, we simply decided that the administrative cost of performance measurement exceeds the benefit of incremental employee productivity. So we don’t do it. Our annual compensation is delivered in the form of Total Compensation Value, which includes any expected target bonus values.
The employee picks their preferred cash and equity splits from this total value. Instead of spending time on performance management processes, we spend our time focusing on setting clear (and high) expectations and delivering regular feedback on our work. In many ways, our system is built on trust. In the typical pay-for-performance environment, top performance is yours to prove and bonuses are yours to gain. At Brex, we pay you for top performance at the onset — but it’s yours to lose if you don’t deliver on expectations. We think this optimizes for consistent, long-term performance instead of short-term, cyclical bursts of output.
Most small startups don’t address (or even talk about) fair pay because the company’s compensation foundations aren’t yet in place.
Most small startups don’t address (or even talk about) fair pay because the company’s compensation foundations aren’t yet in place. But once you have a clean and consistent compensation architecture established, you can now very easily detect and correct pay biases in your system.
Inequities can happen nearly anywhere in the employment process — when sourcing, leveling, creating offers, promoting, etc — so it’s important to watch closely for statistical deviations from expected outcomes based on your racial and gender populations.
Brex runs three pay equity analyses at the end of each pay cycle to ensure we have a fair compensation program. First, for the same job profile with more than 30 people in the role, we use Fisher’s Exact Test to ensure that the gender/race count above and below the median pay resembles our overall employee demographic.
For other roles where we have fewer people in the same job profile and location (which is most of our roles), we measure the average delta from compensation target midpoint to ensure that no gender or ethnicity is receiving unfair compensation treatment. Finally, we review rolling 12-month average promotion rates by race and gender to validate that career advancement is balanced.
Ultimately, we codified each of these compensation principles into the philosophy below. We share this philosophy transparently with current and prospective employees. We think this transparency pays off — in our most recent culture survey, only 1 in 10 thinks our pay is less fair than other companies and we see very little attrition due to compensation concerns. And this isn’t just about the overall pay position. We’re attracting talent to Brex from competitors who offer similar compensation. Our program, however, happens to be more clear, more flexible, more predictable, and more unbiased — which we think is a winning formula.
Brex’s Compensation Philosophy
1. Brex pays at top of market for top talent. Compensation targets per role and geographic market are in the top quartile against our competitors and we use external compensation survey data to inform our decisions.
2. Total Compensation Value (TCV) at Brex is represented in a single dollar value and is inclusive of all traditional elements of annual compensation (new hire equity grant, base salary, bonus, equity refresh, etc).
3. In most roles at Brex, individuals can select their preferred cash/equity split for their TCV. Any subsequent increases in TCV can also be split for cash/equity.
4. Brex makes every effort to offer employee liquidity options for equity during fundraising.
5. We don’t believe in stack ranking employee’s performance in order to distribute bonus funds from a fixed budget. This rewards spot performance and not ongoing improvement through a growth mindset. Instead, our performance process focuses on continuous feedback and ongoing growth.
6. Brex reviews compensation twice per year, in Q1 and Q3 and makes two types of adjustments. Market adjustments are made for everyone in a role if that role’s compensation benchmark has increased or if an employee is compensated under the benchmark range. If an employee is within the compensation benchmark range, a standard increase is made to reflect an individual’s ongoing growth in their role. Brex targets 4–5% annual increases on TCV for standard increases. These adjustments result in continuous compensation growth while at Brex.
7. Brex rewards individuals who have made exceptional contributions to the business with incremental rewards (there are no target distributions or pre-set budgets for these rewards). These employees can be identified through bi-annual values reviews, ongoing monthly feedback, or through management discretion. The Leadership Team reviews all incremental rewards.
8. Individuals whose performance is off track will typically not receive compensation increases.
Fair Pay for Market Labor
9. Brex continuously reviews and adjusts TCV to ensure pay equity across gender and race by geographic market. Brex shares our data with employees on pay equity after each comp cycle.
10. Brex pays local market labor rates for talent. Our compensation target benchmarks reflect talent competition and cost of living by geographic market.
How you compensate your team is obviously an important part of the employee experience, and, consequently a key component of a company’s culture. I hope this gives new startup founders some context on how we approach compensation at Brex and some inspiration to get started building sooner rather than later.