Data Science, Compensation, and Asking for Money

Painting by Josef Wagner-Höhenberg showing a late 1800s farmer signing a contract at a wooden table while a notary gestures to the two bankers at the other end of the table who have a pile of gold and silver coins in front of them.

This post was meant to be a chapter in a book where data scientists who had transitioned from academia to industry shared their experience and advice. Unfortunately, the 2020 pandemic seems to have killed the project, so I am sharing my chapter here. Enjoy!


I didn’t become a physicist for money—quite the opposite, in fact. I remember walking across UC Berkeley’s Sproul Plaza with my father on our way to our favorite lunch spot when I told him, “I love that Cosmology has so few practical applications! It’s amazing that someone will pay me just to expand human knowledge!” Perhaps that should have been my first hint that Physics did not have great long-term prospects, but I don’t think that would have changed my decision to pursue it.

The first time I visited CERN during my PhD, I got an ascetic feeling from looking at the dilapidated buildings and missing storm shutters. Its exterior reflected the same feeling I had expressed to my father a few years earlier—that science was worth more than dollars. I felt instantly at home.

When I ended up leaving academia several years later, it was for stability, not money. I had no intention of dragging my family around the globe just to end up six years older and once again hunting for a job. I’d seen my CERN colleagues try—and universally fail—to land tenure-track positions. So I decided to head back to where I grew up—the Bay Area—and see if I could make it as a data scientist with the help of the Insight Data Science program.

Compensation

As someone who’d never been motivated by money, I discovered during my job search that I was a little naive as to what all the numbers meant. Fortunately, Insight anticipated this and brought experienced data scientists, startup founders, and tech company executives to give us a crash course on life outside of academia. I learned everything there was to know about salary, bonuses, and equity.

I was already familiar with base salary: what you’re paid in semi-weekly paychecks in exchange for doing your job. It’s what I thought of when I heard “This job pays $120,000 a year.” But salary isn’t everything—data scientists often get paid in several different ways.

Every offer I received paid a yearly bonus of 15–20% based on performance. These were from standard tech companies, but I also heard from data scientists in finance and investment who’d be disappointed if their bonus wasn’t an “integer multiple of their salary.”1

Insight also spent a lot of time teaching us about equity: owning a piece of the company you’re working in. For publicly traded companies, things were simple: I would get a certain number of RSUs that would vest over the next three or four years, at which point I could sell them (or hold them).2

Startups were more complicated. I didn’t get any offers from startups, but this is what I learned from Insight and from working with VCs for two years: particularly valuable early hires can receive up to a few percent ownership of the startup, but over time that share would “dilute” as new stock was issued. Complicated vesting structures could mean that even with a good exit, your shares might be worth less than you thought if you’re far down the liquidation preference stack.

After all the preparation, I had two offers to consider. But how to choose? Just pick the biggest number, right? Turns out it isn’t so simple. There are a lot of intangibles—like “what work would I be doing?” and “who would I work with?”—that matter a lot, possibly more than money. I decided to go with the offer that had a bunch of great people on the team and a very interesting problem, even though the total compensation was lower.

Negotiating

Once I had decided on an offer, I entered the most stressful part: negotiating. Insight had drilled into me that regardless of the offer I picked, I had to negotiate it. I spent a few days talking to my friends and family about the offer, trying to work up the courage to call the hiring manager. In the end, I took a deep breath and wrote a quick email:

Hi!

I’m very excited by your offer and the thought of working with the team, but I have another offer for $15,000 more. Can you increase yours? If so, I’ll sign today.

Alex

They wrote back a few hours later, having increased the offer by $7,500. I signed immediately. That five-minute email earned me $18,000 more over two years in increased salary and bonus. A pretty good return on investment for a few minutes of stress.

That time, I had leverage: a higher competing offer. A second offer shows your true market value—someone else is literally willing to pay you that much.

Two years later, I got an amazing offer from my current company, far higher than what I’d said I was looking for during the interview. I knew I had made a slight error giving the first number, and in the back of my mind I could hear my Insight advisor telling me to negotiate. But I was nervous. I did not have a counteroffer for leverage or to fall back on.

I talked to the VCs I worked with for a few days to work up the courage. They all gave the same advice: negotiate. One even said that when he was hired, he was sure his new boss would have seen it as a red flag if he hadn’t tried to get a better deal—after all, making deals was his job. Data science is a little different from investing, but the advice holds: as a hiring manager, I’ve never once minded when a candidate negotiated. I’m just excited to close on someone who’s going to help my team reach our goals.

I wrote an email asking for a lot and got a reply back asking for a phone call. I laid out my rationale—that my friends with similar experience were getting those kinds of offers. They kept me sweating for a few days, but finally returned an offer with a $5,000 higher base salary and $50,000 more in RSUs. It was a lot more stressful than my first negotiating experience because I did not have a fallback option, but again the return on investment was huge!

Helping Others Negotiate

Even if you are not motivated by money, as I was not, if there is some cause you care about—your family, the environment, helping others—you should negotiate. Why? Because money will give you the means to advance your aims. You can donate it, hire someone with it, or put it away for later. In the end, either you or your employer will have the extra dollars; who do you think will put it to better use?

One of my friends got a great offer for a software developer position in the Midwest and was reluctant to negotiate. He asked me, “I have no kids, we own our house. My wife and I don’t need anything. Should I still negotiate?” I encouraged him to do so anyway. With just a few emails, he got a larger signing bonus,3 which, true to his ideals, he donated to charity.

The only leverage he had was that they wanted to hire him and he had not signed their offer letter yet, but that is often enough. Hiring managers spend a lot of time screening candidates; when they find one they think is a good fit, they are anxious to close the deal before someone else does! Offering to sign immediately if your conditions are met is a great way to open a negotiation. You signal to the other party that if they do this one thing for you, they will get what they want: your signature. It’s a win-win.

Another one of my friends got an offer for her first data science job at an autonomous driving company. She had a few other offers pending, but this was her dream job. I had helped her through each part of her job search, and so she called me to talk over the offer. I told her it was a great offer, but since she still had offers pending, she should ask them for $10,000 more base salary to sign immediately.

She initially balked; it was her preferred company, after all, and she didn’t want to lose the offer. I told her they had already spent a lot of money interviewing her and determined that she was the right candidate. Asking for more was not going to jeopardize the offer; they would actually appreciate the opportunity to close on a candidate without having to compete. After a few minutes, she promised to send them an email. A few days later, she called: they’d said yes immediately.

Agreeing to sign immediately and having other offers are two good ways to start your negotiating that I have used and coached my friends in using, but there are many more. I highly recommend reading Patrick McKenzie’s guide to negotiating, where he covers everything from what the employer’s bargaining position is to a step-by-step guide through the whole negotiating process, and even what to say to avoid the dreaded “So what’s your current salary?” question. Patrick’s guide has helped engineers and data scientists earn combined millions of dollars in additional compensation; you should join their ranks!

Finally, remember that money isn’t everything. You can negotiate all the various parts of your compensation, including things you might not consider part of your compensation: vacation days, work location, and the ability to work from home. These can be easier to negotiate because, as an employee, we often value them much higher than the exact monetary value assigned by employers.

Negotiating an offer is stressful but lucrative. Data scientists get many forms of compensation and are in high demand. That gives us leverage. Use it. It’s easier than you think, and you truly have nothing to lose.


  1. Although as a quant friend of mine pointed out: “Zero is also an integer multiple.” 

  2. I generally sell my RSUs immediately; I already have my salary and health insurance tied up in my employment, so selling the RSUs lets me diversify. 

  3. A one-time bonus paid when you start a new job.