Jul 27, 2016

What factors can increase your data scientist salary?

Share:

Albert Einstein was once said to have noted that “compound interest” was the “most powerful force in the world”. It’s sad to say that while most data scientists are intensely data-driven about their projects, they are less so about the compensation which follows that same powerful force of annual growth. Starting low and not taking an optimized view on compensation means that data scientists are often working for less than what they are worth. If you don’t reevaluate, you will always have a sub-optimal data scientist salary.

At Springboard, we’ve been teaching a lot of people the skills they need to become data scientists, data engineers or data analysts. We get a lot of questions about data scientist salaries and different career paths in data science. It’s led us to believe that this is a hot topic. It’s why we built a comprehensive guide to data science interviews–but we wanted to dig even deeper.

Being data-driven people, we wondered what it would look like if you optimized your data scientist career for the highest data scientist salary possible.

We decided to examine several factors that might be relevant. This started with the tools data scientists used and ended with an examination of whether demographic traits determined their career path in data science. We looked into deeper factors like cost of living, but mainly we focused on how to get compensation up as high as possible for a data scientist salary.

We learned that applying to data scientist roles rather than data analyst roles could mean a difference of $50,000 in base salary. We learned that using big data tools like Scala and Spark could add $15,000 to your salary. And we learned that negotiation often leads to a 7% bump in salaries–even though 18% of all people never negotiate their salaries. You should read this guide if you want to maximize your salary as a data scientist.

Do not take this as an indictment of anything that pays less but may be more fulfilling: it is simply a guide to maximizing the salary you should get as a data scientist, and how you can have the most “powerful force in the world” work to your benefit.

How to optimize a data scientist salary

A good starting point for any analysis is O’Reilly’s Data Science Salary Survey. A definitive survey with more than 600 respondents in different industries, it contains different analyses of what traits pay the most for data scientists. The study determined that the mode of base salary for data scientists was about $80,000 USD (assume that for the remainder of this piece we will be using USD for comparison, unless explicitly stated otherwise)–a large amount surely, but perhaps not what one would expect from the “sexiest job of the 21st century.”

Still, quite a few respondents (about 1%) reported a base salary above $240,000, and with entry-level salaries for some companies over $100,000, it’s clear that seniority isn’t the only factor at play here. The possibility of cracking into that upper echelon is tempting for many–but really, what would it take for you to do just that?

Factors

Tools Used

The tools data scientists use are often a point of pride in technical discussions, but they can also make a ton of difference in the compensation you get! We wanted to see what tools gave you the most lift to your average salary–here are the results.

data scientist salary with springboard

O’Reilly’s Data Science Survey singles out the tool that will add the most to your salary: Apache Spark. Scala is another language that is in demand as well. Learning the two could add up to $15,000 to your salary if you assume a causal link in the O’Reilly dataset (a dubious assumption, but one that shows a strong correlation at least).

Learning D3, a Javascript visualization library, is the only significant data visualization tool to have an effect on your salary with a correlation of a $8,000 positive boost.

Lastly, familiarity and experience with cloud computing will also boost salaries, with respondents who use Amazon Elastic Mapreduce getting a boost of about $6,000 in their salaries.

O’Reilly also noted that those who used 15 or more tools could make up to $30,000 more than those who just used 10 to 14. There is clearly a premium associated with mastering a variety of tools.

The main takeaway from the O’Reilly Study is that there are nine main clusters of tools that data scientists use in their day-to-day, ranging from the Hadoop ecosystem to the open source environment surrounding Python, to the closed Microsoft SQL cluster. People tend to learn tools within a cluster, as the tools complement one another within. People who tend towards closed Oracle and Microsoft tools will earn less, while those who flock to open source clusters will tend to earn more.

It should be noted that O’Reilly’s audience tilts towards one of open source practitioners, as the O’Reilly ecosystem is focused on supporting and highlighting new technologies. With that said, these results tend to hold up pretty well.

Searching in San Francisco in Indeed with the keywords Python and Data Scientist yielded an average of 31% of expected salaries being below $105,000 while searches with the keywords Scala and Data Scientist had no expected salary below $105,000.

How to get the best data scientist salary possible based on the tools you use:

  1. Learn Spark and Scala
  2. Learn D3
  3. Get comfortable with cloud computing, especially Amazon Web Services
  4. Get into open source tools and stay away from proprietary ones
  5. Learn a whole lot of tools (aim for 15+)

Industry

Where you choose to work can determine your salary in many ways. Different industries have different data challenges, and different abilities to pay for top data science salary. Make sure you’re not constraining yourself by working for a company where you can’t capture your full value.

data scientist salary with springboard

Figures are approximate

The O’Reilly Salary Survey found that the highest salaries went to data scientists in search/social networking, which makes sense given the amount of valuable data those kinds of companies sit on (think LinkedIn, Facebook or Google). They track the interactions of millions of people on their platforms and must come up with some reasonable conclusions based on that chaos.

data scientist salary with springboard

The Facebook data scientist salary is an average of ~$133,000 a year based on 44 employee salaries. Google has a data scientist salary around ~$145,000 a year

Payscale takes technology companies like Amazon that pay about $150,000 in median data scientist salaries and compares and contrasts that with consulting companies like Booz, Allen and Hamilton, whose median salary for data scientists is below $100,000.  Senior data scientists at Linkedin earn an average of ~$157,000 a year based on 26 employee salaries.

It’s important to note too that companies like Facebook and LinkedIn also offer generous stock incentive bonuses, which easily add about $40,000 to $50,000 more when it comes to compensation. Those companies continue to hire, with Facebook having 16 open positions in data science.  

Startups are often big hirers as well and indicative of where salaries are headed in the future. While there were only three data scientist jobs from hardware startups that paid over $100,000, 71 enterprise software startups were looking to hire data scientists at a salary above $100,000, perhaps indicating a latent demand for data scientists in the industry that will soon propagate to larger companies.

How to get the best data scientist salary possible based on industry:

  1. Look, above all else, to work in social media or search companies like Google or Facebook
  2. Keep an eye on trends in startup salaries for data science as they may spread to industry
  3. Look towards hardware, finance, and software companies if you don’t want to work in social media or search companies.

Location

There is significant advantage to working in certain areas with a concentration of top talent and thriving companies. The O’Reilly Salary Survey reports that the United States of America has the highest median salary and range for data scientists. Now it’s up to us to find the most lucrative area in the country.

Most people think of Silicon Valley as the obvious pick, and at first that seems to hold up. Payscale reports an uplift of 23% for Mountain View where Google and LinkedIn are headquartered. The O’Reilly report even managed to infer that living in California was worth an extra $16,000 in salary.

data scientist salary with springboard 

That isn’t the whole story, though. We haven’t taken into account cost of living or state tax rates.

We’ve listed the top-paying areas in the United States when it comes to data scientist salary and compared them to cost of living indexes for each city so you can get the complete picture (for more on our methodology, see our appendix at the end of the article). We then subtracted the state tax rate for the median salary for a data scientist ($105,000) and got the following chart.

Salary minus Cost of Living minus State Tax (Rounded).png

Some key insights on this:

  1. San Francisco is not actually where you can make the most as a data scientist: neighbouring San Jose takes the cake with lower cost of living and higher salary differential. In fact, by adjusting for cost of living and state taxes, Seattle actually comes out over San Francisco!
  2. Cities like Los Angeles and Austin seem like bad deals when you consider how much lower the base salary is for data scientists, but in reality their low costs of living and in Austin’s case, no state taxes, make them comparable to other tech hubs.
  3. Stick to the West Coast over the East Coast when it comes to data science salaries.  

How to get the best data scientist salary possible based on location:

  1. Look to work in the United States
  2. Look to the West Coast of the United States
  3. Consider cost of living and state taxes: optimize for San Jose

Roles

A lot of people conflate data scientists with data analysts and data engineers, a mistake that led us to write extensively about the differences between the roles.

Data Science Roles and their Average Salary + (Approx.) Range (1).png

Ranges based on those sourced at this primer by Datajobs, averages sourced from Indeed. It should be noted that the ranges taken are sourced as a combination of entry level and experienced ranges for data analysts and data engineers.

Data scientists have a much higher salary ceiling than the other two roles, especially those entering data analyst roles.

Roles in practice: How can you differentiate between the roles? Most of the time it will be stated in the job title, but if you really want to parse out exactly what kind of role the company has in mind, look to the skills they expect and the work they’ll look for you to do.

If Excel and SQL are strong requirements and the job description sounds like you are going to be querying data often but not building data systems, then you’re applying for a data analyst role. If programming knowledge is a strong requirement but knowledge of statistics and algorithms is secondary, you’re likely applying for a data engineer role. If you’re being asked to manage data science projects from end-to-end and to implement your knowledge of business, communication, algorithms, and programming, it’s likely you’re applying for a data scientist role.

To maximize data scientist salary select the data science role.

How to get the best data scientist salary possible based on the role you choose:

  1. Make sure you’re applying to data science roles, and not data engineering or data analyst roles in disguise.

Experience and Degrees

Increase in Annual Salary vs Years of Experience or Degree.png

In O’Reilly’s survey, adding a Master’s degree without a Ph.D is correlated with about a $1,000 a year increase in data scientist salary. Adding a Ph.D degree is correlated with $9,000 a year in increased data scientist salary.

Payscale indicates 5-10 years of experience is worth ~$20,000 more in annual salary, 10-20 years of experience is worth ~$30,000, and more than 20 years of experience is worth ~$55,000.

A search for data scientist in San Francisco yields that 42% of data scientists earn a salary above $100,000–while 68% of senior data scientists earn above $100,000. Experience matters for data science, perhaps even more than different degrees.

You’ll likely need an advanced technical degree to get your foot in the door:  

However, a Ph.D can quickly become a massive investment of time (with an average time of 8.2 years from start to dissertation), and in that time, if you aimed for at least five years of experience in data science, you’d get double the return on annual salary. Still, you can’t maximize your data scientist salary without a Ph.D.

How to get the best data scientist salary possible based on experience and degrees:

  1. Get a Ph.D. or a Master’s that gets your foot in the door
  2. Aim for many years of experience in data science roles

Negotiation

When you’re presented an initial offer, the employer always has buffer room with which they concede some space for negotiation. Your initial salary will quickly become your anchor for future salaries, so you’ll want to make sure you negotiate well in order to maximize your salary. Those who rated their negotiation skills in the O’Reilly Salary Survey as being very poor had a disadvantage ranging from $25,000 to $50,000 compared to their peers who claimed their negotiation skills were excellent. 

An astonishing 18% of people never negotiate their salary, despite the fact that those who do typically see their salary raised by 7%.

You can develop successful negotiation skills even if you’re an introvert. Don’t hesitate to do so.

How to get the best data scientist salary possible based on your negotiation skills:

  1. Practice negotiation
  2. Always negotiate

* Demographic Traits

This last section is dedicated to unearthing an uncomfortable reality when it comes to data science salaries: data science, like any career, has some biases when it comes to demographic traits. The O’Reilly Salary Survey indicates that women earn $8,000 less than men even after adjusting for company size and location.

There is a contentious debate over whether or not there is a gender gap in technology. There is a noted debate about diversity in technology across racial and gender lines. This affects who gets hired and who gets the best salaries.

This bias is not something that can be easily corrected individually–it is something to consider systematically. You’ll have to know that many of the same forces that create biases in hiring in different industries will be present in data science as well.    

Conclusion

We applied data science to data scientist salaries and came up with a bunch of actionable insights. They are summarized below.

722e4832-f598-45c2-9579-1794f80f0a53.png

You will need to follow these steps, enumerated above, to harness the “most powerful force in the universe”–compound growth–and get the highest data scientist salary possible.

Appendix

 

City

Data scientist salary difference from national average

Cost of Living Index (based on NYC)

State Tax Rate (income of $115,000)

Salary minus Cost of Living minus State Tax (Rounded)

San Francisco

+22% (1.22)

0.97

10.30% (0.103)

147

New York City

+5% (1.05)

1

6.85% (0.0685)

-11

Seattle

+6% (1.06)

0.9071

0%

153

San Jose

+26% (1.26)

0.8587

10.30% (0.103)

298

San Diego

0% (1)

0.7733

10.30% (0.103)

124

Boston

-3% (0.97)

0.8937

5.10% (0.0510)

25

Los Angeles

-7% (0.93)

0.6941

10.30% (0.103)

133

Austin

-8% (0.92)

0.7826

0%

137

Our cost of living index comes from Numbeo and is based on prices in New York City (which is the index 1 and the highest-priced place to live in the mainland United States). We sourced data on the salary difference from the national average on Payscale. We’ve subtracted the cost of living and state taxes on an income of $115,000 (a median salary for a data scientist) to get the total differential.

Get a data science job

Get a data science job