Compensation Restructuring (Post Series A)
Hi! I’m John! I’m a co-Founder of a Series-A company which recently restructured compensation for a Series A through Series B transition. This article is useful for any founder who is building out their compensation scheme. If you are interested in HR, or building novel systems in cryptocurrency — please reach out, we are actively hiring!
I’m incredibly grateful to: Tity Lyngdoh, Sutton Kauss (if you have HR questions, please reach out to her!), Rafael Lopez, Katherine Markel, Justin Lee, Michelle Luong, Kevin March, and Van Phu for all their advice and assistance with this.
Table of Contents:
- Summary
- Motivation
- Process of Updating Compensation
- New Compensation Scheme & Model
- Conclusion
Summary
Reasons for doing it
Our company is about to undergo a sizable scale-up. As of 3/6/22, we have 30 full-fledged floaters. Our plan is to target 200 full-fledged floaters by the end of 2022. This represents about a 600% increase in personnel. It strained us to maintain consistency and fairness in our compensation for the first 30 and we dreaded what it would look like at 300. So we cleared the HR debt and systematized it.
Compensation is something that is incredibly personal and speaks to what our company values. A startup’s only true assets are the people that are a part of it and a company’s objective should be to both achieve its external mission and to nurture the people inside of it. Learning, Happiness, and Compensation are the three things that an organization can give to its members, and it should give those in abundance.
What you will get from this
This post describes a learning process about compensation, the ultimate conclusions that we discovered, and how we are now using those for the next chapter of the company.
Motivation — how did we do it before
How we compensated sub 30 people
We used three sources:
- Aeqium’s compensation tool — used to price equity
- VCECS — good for high level ideas for senior folk
- Buffer’s salary calculator. If you know nothing on compensation, Buffer’s salary calculator is probably the best 80/20 you can achieve.
Our initial method of benchmarking was coarse. We had roughly two ways of getting industry salaries:
- Take the title of the job, and go to Glassdoor, Levels.fyi, and other locations to estimate industry pay. Take the average, and add 20% (pay above market rate) to get the Total Compensation Value.
- Split the organization into two tracts: Business and Engineering. For the engineering team, base it off Levels.fyi. We had 3 levels we then placed engineers in: L1, L2, and L3. These corresponded approximately to L3, L4, and L5 at Google respectively. For the business team, we used our best definition of
L1in the business world and used the same ratios/experiences to create business bands.
This system had several issues:
- We treated business roles equally because it was easier than creating a ladder for each non-engineering speciality.
- We had little understanding of how the industry was benchmarking people. We couldn’t accurately understand how much the person was previously making/should be making.
- Our internal leveling was ad-hoc and the benchmarks for those levels (outside of engineering) were quite poor.
- Negotiating was becoming a substantial issue. We didn’t have clear guidelines on what we could or couldn’t do for certain roles and it created tension in the interview process.
- A lot of people asked often when we review compensation, how we think about, and how we benchmark it. We had no clear answers.
Because of these reasons, we realized that we needed to update our compensation.
Process
What we learned
During these conversations, we did three things:
- Wrote down the original process that we use to compute compensation. It was a lot easier to modify a process and explain its shortcomings if you write down what the process is.
- Drafted our compensation philosophy. Just tenets that we thought would be good to be guided by.
- Talked to people about how to price equity.
The draft compensation philosophy was pretty useless itself, but instead discussing what tradeoffs it promotes made it highly valuable. For instance, the value of we pay people based on the value they provide is, I think, mostly agreeable. However, reframing that as:
- We pay people in NY the same as those in India
- We do not incentivize people to live near the office (indeed based on net pay/effective rates, we discourage it)
- We do not pay experienced people at the company more than others joining recently
Gave a lot more stimulating conversation on whether we wanted that to be a core part of our compensation philosophy or not.
It took us a while to find good data sources. Finally, we discovered enough sources — it was hard to find banking (McLagan) and regulatory data (OptionImpact and Radford), and then getting a more clear picture on SWE (OptionImpact, Levels.fyi) was difficult as well.
We mapped out departments to different data sources:
New Compensation Structure
There were five deliverables for the end of the project:
- Compensation Philosophy
- Compensation FAQ
- Benchmarking
- Equity Pricing
- Calculation worksheet for compensation
Compensation Philosophy
We settled on a five part compensation philosophy.
Compensation FAQ
We put together a page summarizing how Compensation works, how we calculate the Total Compensation Value, and answering some common FAQ questions on it. The idea is that this is shared with all employees of FPG — we aren’t quite comfortable yet to share this with all applicants, but it is a near term goal to achieve that level of transparency.
We also created an FAQ document with all the questions we have been asked:
Benchmarking
Data Sources
We use 4 data sources:
- OptionImpact — free, great data for startups with equity and salary breakout.
- Levels.fyi — free, great data for engineering roles and organizations. Has most big tech in it
- Radford — cost. Good general purpose data (has Janitors and CTOs), but more designed for SME businesses
- McLagan — cost. Has investment banking data, which nearly no one has.
As auxiliary, we use a few others:
- Pave.com — free, good at auto-generating levels and seeing them. They use the same leveling system as Radford, and worked with an outside firm on creating the underlying model. Their data comes from everyone on their system, and the system is free.
- Carta total compensation — cost. Will auto-pull from Carta & your payroll provider to show where you sit with payments and salary. Newer tool, but pretty cool.
It is important to note, if you are doing this as a startup I would strongly recommend OptionImpact & Levels.fyi. Under few conditions would you need to venture out from them. Carta’s tool is a bit early, but honestly looks pretty promising to reduce some of the searching and have quick integrations.
Equity Pricing
We created a page on equity pricing.
Equity pricing is used for two things — to inform candidates how much their offers/options could potentially be wroth, and to inform how many shares should be given for certain levels of compensation.
We struggled with how to communicate this, but eventually decided that the way it was done in the Recruiting book was probably the best method. Showing just the range of possibilities:
Most companies use internally either best value (which is last priced round valuation), or some multiple of their revenue/spot. We used a bit of a conservative average — slightly above our best value as we have grown quite a bit since out last round.
If you are a start-up post Series A, using best value may be a good 80/20 to get consistency to other organizations.
Contributions
Compensation is a challenging subject. We were very fortunate to have an amazing team and group of colleagues who embarked on this project with us, but for many people that don’t have those resources — please reach out and let us know what edits/items we should do better!
