Startups Employees Perks & Incentives, part1: Wages
There are two main components of a compensation policy: salaries and equity.
An equation with only two variables? Should be pretty simple, right? Well, not when you are talking about something as symbolic as money.
Before all, let’s remind the core principles of any good compensation policy:
- Being as objective as possible: this ensures fairness and acknowledges a basic truth: people talk. The goal is to ensure fairness, real and perceived (more or less the same package for the same position, all things being equal),
- Cash is for a short/mid term reward when equity is a long term alignment,
- Compensation needs to be adapted to market practices (especially, to local practices).
That’s being said, let’s dig dive and look at the best practices of a compensation policy for startups:
How To Find The Right Wages For Your Employee?
1. Theory Will Lead You to Market Practice
In the economic theory, wages are linked to the value created by the employee. The problem is that value creation is very hard to measure. Sometimes someone’s productivity is simple to assess (eg: if you manufacture products all by yourself). But once you start to put in place some kind of division of labour, it starts to get very complex: different people contribute in different ways to the company’s value creation. It is not rare to see a team overall productivity fall when someone considered as a “low performer” leaves. Team players have all some significant effects in business as in sport: a living organism has very complex mechanisms with many feedback loops.
Thus, using pay as a proxy for value, the question is “what is a maximally fair assessment” or put another way “what is the least biased distribution of attributing value?”
Venkat Venkatasubramanian, Samuel Ruben–Peter G. Viele Professor of Engineering at Columbia University and author of How Much Inequality Is Fair?, answer to this question that way: “In a competitive free market environment, companies and employees as rational agents arrive at this distribution iteratively, by a trial-and-error evolutionary feedback process, through the free exchange of information and people between companies and the market environment, until equilibrium is reached. The survival instinct drives people to maximize their values and trade their current jobs for more rewarding ones. Companies also do the same by hiring and firing employees in order to derive better value from them and maximize profits. Therefore, the least biased distribution of relative value is reached in practice via such a market process, empirically. That is, the equilibrium distribution is ―discovered through such evolutionary adaptation”. (Venkatasubramanian, 2009)
Said another way: the right compensation is market compensation.
So the best way is to start by taking the time to make benchmarks and figuring out the right ranges for the different employees' positions.
2. Establishing the fixed salary
The best way to establish the base salary is to look at a benchmark of comparables company.
Yet, you cannot just rely on benchmarks since you won’t have the time to have continuous and granular benchmarks for every position in your company. Jean-Charles Samuelian, co-founder of Alan, shares two tips on that matter: (1) you don’t need to focus on a comprehensive benchmark, in reality, if you are competitive to 4–5 companies that are either direct competitors or very attractive startups, you are alright; (2) if you start to have a gap between your compensation and the market, you’ll discover it quickly in your next hiring process through your candidates feedbacks and claims.
Even though, you cannot rely exclusively on benchmarks and market practices, so you will have to build grids of compensation across your company. For each type of position (“tracks”), you should have a dedicated grid. You’ll end up with grids for sales, developers, product managers, etc. The bigger your company, the more tracks you will have (eg: data scientist, buyers, account managers, etc.). These grids should reflect market compensation, becoming efficient proxies between two compensation benchmarks — wages don’t evolve drastically every year, especially in places where there is no shortage for talents. Grids ensure fairness across the company, offer visibility for your employee and give you some latitude: you can offer some premium or discount to market salaries. For instance, the first year or so, you can attract great people with a salary below the market if you give high enough equity package. On the contrary, if you are a brand-driven company, you may choose to pay your brand marketer above the market practice to make sure they won’t leave.
3. Designing Your Grids: From Benchmarks to Salary Formula
Buffer, being a distributed team across the world, had to make sure they were competitive with the local market practices and keeping a fair formula for all their employees.
To build their grids, they designed a transparent salary formula, which has the benefit of making the salary readable and objective (let alone the experience which is discretionary): Role * Experience * Loyalty * Choice
Role, being composed of the 4 following variables”
- Overall base: they use US data from Payscale and Glassdoor as their benchmark for 35% of the base
- Location base: they factor location’s cost of living using Numbeo together with data from Payscale and Glassdoor, for the other 65% of the base
- Cost of living correction: a premium of $0-$8,000 yearly to adapt the local cost of living.
- Role value: last multiplier to adjust the overall salary, because they don’t “agree with the market salary data all the time (for customer service roles, for example) and so [they] create our own “role value adjustment” based off what [they] feel is fair.”
Experience: they weight the role base with an experience multiplier:
- Beginner: 1x
- Intermediate: 1.1x
- Advanced: 1.2x
- Master: 1.3x
Note: it’s good to have between 4 and 7 levels of seniority, and make sure that the multipliers will match the market practices. The multipliers might be different between the positions and the increase between the levels might not grow linearly.
Loyalty: raise of 5% every year for everyone.
Note: salaries should grow every year at least more than the inflation rate.
Choice: they offer every member the choice of having an extra $10k to the salary or get roughly 30% more stock options;
Here is an illustration of their formula for an advanced engineer, living in Cape Town, who chooses more equity: $60,662 x 1.20 + $9,000 + $0 = $81,794
Buffer’s approach to compensation policy implies very clear and simple algorithms, which ensure fairness and lead to precise compensation.
4. Using A Universal Grid
An alternative way of building your grid is using Alan’s method, which is very effective, by being simple to communicate and use. The grid is generic to all position: lines are a function of the position, its scope and the level of expertise; the columns are the years of experience.
To use it, you take a certain position (eg: you want to hire a new engineer) and look at market practices. You determine the range where the candidate can be located in the grid: an ENG of that seniority (eg: between 1 and 7 years) will be in between the lines D to H for the compensation to match. Then you just need to figure out what are the attributes for the different lines for an ENG (more about that in the chapter about performance review), and the discussion about salary will only be about figuring out what the scope and expertise are (the attributes of this candidate for that position) to know which line (s)he should be.
5. Salaries Need To Be Fair, Which Does Not Always Mean Equal
Zoe Cullen and Ricardo Perez-Truglia, in their paper “How Much Does Your Boss Make? The Effects of Salary Comparisons” (2018), studied how employees learn about the salaries of their peers and managers and how their beliefs about those salaries affect their own behaviour.
They discovered two very interesting findings.
First, employees are very impacted by differences of salary within their peer groups. “Perceived peer and manager salaries have a significant causal effect on employee behaviour. […] While higher perceived peer salary decreases effort, output, and retention, higher perceived manager salary has a positive effect on those same outcomes.”
Second, this effect does not happen when employees learn that their managers and boss earn even significantly more money than they do. “Even though employees are discouraged when they find that their peers earn more than they thought, employees, are not discouraged when they find out that their managers earn more than they thought, even when they do not expect to reach that managerial position themselves.“ The last part is interesting: you could think that differences of compensation might be accepted by employees as they could one day reach that package. Yet, it is not the case. This incentive does work, yet it is not detrimental even for the one that won’t be able to go up in the ladder.
The key is to think in term of fairness (same salary for the same responsibility, and never forgetting that employees care mostly about their standing in a specific reference group — their peers), which does not hinder you from using money as an incentive to get a promotion.
Thus: (1) you should have the same fixed salary within a peer group, (2) you can put in place differences of compensation within the vertical ladder, (3) the only differences of compensation within the peer groups must be fair and readable, so it should be put in place through a variable salary.
6. Variable Salaries: Aligning, Boosting and Rewarding actual outcome
The main trade-off regarding wages is the split between fixed and variable salaries. The rule of thumb is to give 20% of the fixed salary in variable. In the US it is on average 50–60% and can go up to 100% in certain companies.
A good variable policy is a management lever to (1) link individual contributions to the overall strategy of the company, (2) catalyse everyone’s outcome (by increasing motivation), (3) fill the gap between market practices and the actual outcome.
Some high performers feel their contribution is not well rewarded. Variables should mitigate that risk of having a decorrelation between productivity and compensation. Indeed, there is a tacit agreement in the job market: the cost of living needs to be taken into consideration. And generally speaking, the older you are, the higher is your lifestyle, which end up with a premium for age. So the correlation, all things being equal, tend to be more between age and salary than productivity and salary. This divergence is not always a problem. It is important that companies ensure the actual needs of their employees: it matters for economic reasons in the long term (satisfaction, retention, attractivity, etc.) but also because companies should answer to broader rules of fairness and social justice.
The point is that a good talent management policy will not only align and maximize the value creation, it should be flexible enough to attract and retains outliers (high performers and talents in shortage).
Note: Beware of signing bonuses. It is often used as a way to attract key people that are highly paid without having to factor a recurring high salary. Yet, if the next year’s salary doesn’t match the first year’s package, it will be lived by the employee as a decrease of salary and you might lose your talent. And if in the end your growth allow you to pay the same salary, you better just offer the real package from scratch, without the signing bonus.
7. Establishing the variable salary
The composition of the variable is a bit tougher to establish. And variable are not compulsory. For some positions, it is required (eg: sales), for others, it might not be relevant (eg: data scientist). You need to keep in mind that variables imply to establish and follow good performance review processes. So avoid putting a variable in place when you don’t feel it creates real benefits (low impact on motivation, no gap with market practice, low risk of producing a sense of injustice).
For the position that will have a variable salary, you’ll have to set up the parameters.
First of all, you have to decide whether you cap it or uncap it. When most of the objectives are personal and/or the variable formula is directly link to the generation of EBITDA, the best solution is often to uncap the bonus. Sometimes you’d like to control the OPEX, no matter what growth you generate.
Secondly, you need to figure out how often you pay the variable. The more often you pay it, the more attractive it is (the reward is never far away). You can even pay retroactive rewards in case of catch up (eg: 90% of the objective of Q1 but 130% in Q2), especially if you chose to cap the bonus.
Thirdly, and maybe more important, you need to find a good and precise formula for the attribution of the bonus for each one of the employee. It is always a good idea to have around half the compensation depending on a collective outcome (eg: for the company scale, reaching a certain level of gross margin and/or EBITDA ; for a team, managing to develop a new feature or to develop and launch a new inbound strategy) and the other half as a specific functional goal (eg: acquisition of 10000 new users).
Beware of not falling in the trap of trying to make everything quantifiable. Some tasks/activities are qualitative and the bonus attribution needs to be discretionary. As long as the decision-making process is transparent and fair, this is not a problem. It that case, it is always a nice idea to identify and explain precisely the expectations and the things that will be used to assess the performance.
8. Salaries should be updated as often as needed
Fred Wilson tells “Figure out what “market salaries” are for all the positions in your company and always be sure you are paying “market” or ideally above market for your employees. And review your team’s compensation regularly and give out raises regularly. This stuff matters a lot. Most everyone is financially motivated at some level and if you don’t show an interest in your team’s compensation, they won’t share an interest in yours (which is tied to the success of your company).”
Nextflix, has made their compensation philosophy famous thanks to their amazing culture deck (slides 95–111): pay top of market, because one outstanding employee gets more done and costs less than two adequate employees. This requires a frequent update since market compensation change often.
Netflix has three tests that they use at the time of hiring and every following year since market practices need to always be updated to be effective:
Test 1: What could this person get elsewhere?
Underlying principle 1: pay them more than anyone else likely would
Test 2: What would we pay for a replacement?
Underlying principle: Pay them as much as we would pay to keep them if they had a higher offer
Test 3: What would we pay to keep that person?
Underlying pay them as much as a replacement would cost
Note: Your company might not be able to afford paying all your employee top of market -this happens often when your market is competitive with big players paying well. You may chose in that case to pay top of market the people filling the positions most essential for your growth.
9. What about founders salaries for VC-backed startups?
Establishing the salaries of founders is never easy. Their profiles are very eclectic and for founders, the real incentive should be equity, not salary.
The general principle is the following: being able to earn more money in the long run if the company is successful thanks to equity, but having a discount in the present time (especially before market-fit), compared to what they could be earning in an established company. Consequently, the salary will be below market practices for a given profile of founders, so the needs of the founders have to to be taken into consideration. As Fred Wilson wrote, “people can’t use options to pay their rent/mortgage, send their kids to school, and go on a summer vacation with the family.” Indeed, salaries won’t be the same between a young founder, a CEO that sold his company and founders having children and living in a big city. Also, you have to balance the total revenue stream and the actual cost of living (eg: in France, there are many founders that are benefiting from the unemployment office).
Bottom line, for founders, salaries should never be a source of additional stress and respect the general principle that equity should be the main driver of their incentive. Market practices in France: in the seed stage founders are paid (annual gross salary) between €40 and €80k (more or less a Gaussian distribution centered in €60k plus many founders at 0 when benefiting from the unemployment office), in series A between €60k and €100k (more or less a Gaussian distribution centered in €80k). A comprehensive overview of the market practice in the US can be found here.
That’s all for the moment! You should have a good overview of how to design your compensation policy. In the part 2, we discuss how to determine what level of equity should be offered to a given candidate. Stay tuned!
Questions to Ask Yourself
What is the philosophy behind your compensation policy?
Is the compensation fair, effective and well communicated?
Do you have benchmarks (industry/peer group) to calibrate the salary?Should there be a variable salary? Why? How much?
Reading Further
This article is an extract from the book Human Resources For Startups, that you can find in your local Amazon (US, UK, GER, FR, ES, JP, IT!)
Let’s Keep In Touch
This article has also been published in my weekly newsletter OneStepFurther. You can subscribe here to receive the next edition.Recommended Readings
Compensation, George T. Milkovich, 2016
Strategic Compensation, Joseph Martocchio, 2015
############################################
Special thanks to Mathieu Daix, Marie Ekeland, Pierre-Yves Meerschman for their insights and Déborah RIPPOL for her feedbacks!
