The world’s leading publication for data science, AI, and ML professionals.

How Much Do Data Scientists Make Part 2

We Find Some Surprising Trends When We Dig Deeper

Photo by Lukas from Pexels
Photo by Lukas from Pexels

Last time we used H1B salary data to explore general data scientist salary trends (please refer to my previous blog post for details on the data). Here is a quick refresher of what we found:

  • We found the median data scientist base salary to be $120,000 (not including bonuses and equity).
  • There was very little year over year growth in the median salary.
  • The biggest tech companies generally pay the most.
Median Data Scientist Salary by Year
Median Data Scientist Salary by Year

Digging Deeper Reveals 2 Underlying Trends

Last time I didn’t have enough time to dig more deeply into the numbers. But after examining things in more detail, I discovered some surprising trends.

First, the general stagnation in Data Science salaries on average over the past few years is really a tale of 2 trends:

  1. Within the "trendier" tech companies (household names like Google, Facebook, Apple, and Airbnb as well as smaller firms like Stitch Fix and Opendoor), data scientist salaries are growing substantially.
  2. For the other companies in my sample (firms like IBM, Cisco, Oracle, Boeing, Nordstrom, Schlumberger, etc.), median salaries have actually slightly declines.

I realize that my list of "trendy" tech companies is completely subjective – my criteria was, "Would the average grad of a top engineering school be excited to work there?" Regardless, check out the difference in trends:

Trendy tech raising data science pay, everyone else not so much
Trendy tech raising data science pay, everyone else not so much

My guess is that there are two groups of companies – ones that have the ability to produce, gather, and extract competitive edge from proprietary data and ones that do not have it. Because the "haves" can get more value out of their data scientists, they are willing to pay them more as well. Meanwhile the "have nots" are much more likely to get frustrated with the lack of value created by their data science teams resulting in a lower perceived worth (and higher turnover).

Companies like Google and Facebook have demonstrated that gathering and exploiting unique data can create a massive competitive advantage (and this competitive advantage expands over time as more data is collected and insights are gleaned). Their success with using data science to generate massive profits creates a virtuous feedback loop where data and research creates profits which are then reinvested back into more data and research (resulting in even greater future profits).

So the lesson for aspiring data scientists is to work for a company that owns the type of unique data that grants a competitive edge. And a company that knows how to properly use that data to create a virtuous loop of research and profits.

Here is a recent (2018–19) look at median data scientist salary by company. Remember, we are looking at only the data scientist job title and I filtered out companies that hired 5 or fewer data scientists during the period.

Data scientist salary by company, those Stitch Fixers make a lot!
Data scientist salary by company, those Stitch Fixers make a lot!

Wow, look at Stitch Fix! That is a pretty penny especially when we remember that the $190,000 median figure represents just base salary and does not include bonuses or stock. More generally, the companies at the top are all data driven and either have massive amounts of proprietary data (AirBnB, Lyft, Facebook) or are building towards it through unique interactions with their users (Stitch Fix, Opendoor).

Lastly, when looking at these differences across companies, it is helpful to remember that they are driven not just by the relative sizes of each company’s bank account but also by differences in role and required expertise.


Silicon Valley Data Science Pay is Higher (But We Still Can’t Afford the Houses Here)

Unsurprisingly, data scientist pay in the San Francisco Bay Area (Silicon Valley) is higher than in other parts of the country (many of the companies in my "trendy" tech group are located in Silicon Valley). The difference in median base salary shown in the chart below is probably understated as I focused primarily on the West Coast and other tech hubs – many low cost cities in the Midwest and the South were left out of my analysis.

Silicon Valley firms (orange) pay more
Silicon Valley firms (orange) pay more

The difference in salaries between data scientists working in Silicon Valley versus other regions looks pretty significant. But unfortunately, the differences in cost of living (driven by sky high home prices and rents) is even greater.


So Where Should I Go to Live Most Comfortably as a Data Scientist?

Making big bucks is great but we need to remember that expenses matter just as much. If we make big bucks but then we end up paying even bigger bucks to Uncle Sam (taxes) and our landlord (rent), then we may still end up a net loser.

First, let’s take a look at how data scientist salaries stack up by region (2018–19 salaries only):

Data scientists earn more in major coastal cities
Data scientists earn more in major coastal cities

As expected, the San Francisco Bay Area comes out on top in terms of median base salary. New York salaries are less than I expected them to be – the likely reason is that my dataset is not capturing some of the data scientists employed by the big banks (where they are probably called quantitative strategists, risk managers, or portfolio consultants). Also, using just base salary probably understates the earning power of data scientists employed by the big banks as their compensation is weighted heavily towards annual cash bonuses (100% bonuses are not out of the ordinary).

The one that really jumps out to me is Seattle. As a data scientist there, you are earning almost as much as you would in Silicon Valley, but you pay no state income taxes and enjoy significantly lower housing costs.

Let’s divide the median base salaries from the previous bar chart by median rents (multiplied by 12 to get annual rent) in each of those cities to get a rough proxy for how far your dollar goes (I left out the "Other" category as that includes multiple cities, but not enough to safely apply a national average). In case you are curious, here are the monthly median rents that I used (sourced from this article):

  • Bay Area (Silicon Valley): $3,300, a bit lower than the value in the article to reflect slightly lower rents outside of San Francisco.
  • Seattle: $2,100
  • New York City: $2,700
  • Los Angeles: $2,400

And here are how the base salaries look when converted to units of rent-equivalents. Recall that I am taking the monthly rents and multiplying them by 12 in order to annualize them:

Your dollar goes furthest in Seattle
Your dollar goes furthest in Seattle

Adjusted for rent, your data science salary goes a lot further in Seattle versus other tech hubs. And this is before we account for the fact that you pay lower taxes (no state income tax remember?) as well. So if you don’t mind the rain, consider dropping a resume or two with Amazon, Microsoft, and Starbucks.

Finally, let’s check how the tech vs. no tech salary disparity looks across these regions (2018–19 salaries only). Seattle "trendy" tech companies on average pay more as well, just like the Bay Area (Silicon Valley). The disparity in New York is especially large – my best guess is that non-tech NYC firms pay large annual bonuses, which would not show up in the H1B salary data.

In all regions, trendier tech companies pay more
In all regions, trendier tech companies pay more

Conclusion

Before we call it a day, let’s recap some of the key findings from our data:

  • There are two underlying trends in the data scientist salary data – steady growth in data science salaries at "trendy" tech companies and stagnation in salaries paid at other firms (though the median salary among these companies is still close to $110,000).
  • Looking at current data (2018–19), there is a significant spread (more than $30,000) between what you can earn working at a "trendy" tech company versus other companies. And this spread is just base salary too – the actual disparity is probably even bigger after accounting for bonuses, equity, and the performance of that equity (trendier companies have trendier stock).
  • Seattle offers very competitive data science salaries, no state income tax, and a relatively reasonable cost of living (in terms of median rent). And you get the opportunity to build character by toughing out the rain and cold.

Hopefully this was insightful and useful to you. Cheers!


Read Part 1 of This Story Here:

How Much Do Data Scientists Make?


More Data Science and Analytics Related Posts By Me:

What Do Data Scientists Do?

Are Data Scientists at Risk of Automation

Got Data Science Jobs?

_Understanding PCA_

_Understanding Random Forest_

_Understanding Neural Nets_


Related Articles