First the plug: IQ Workforce is the leading recruiting firm for the analytics community. We provide full-time and contract analytics talent to clients globally.
Our team speaks with analytics professionals every day. We gather a tremendous amount of career data: salary, skills, experience, motivation, fit factors, etc… so we are in a good position to set the bar when it comes to analytics compensation.
This is not the first time that we have reported our proprietary salary data, but there are some significant changes this time.
Firstly, we decided to eliminate titles and focus instead on years of experience. We have found titles to be less and less meaningful (particularly when discussing digital analytics professionals) because there is widespread title inflation in the space. Analysts with 2-years of experience in some companies are called Managers. Directors in many companies don’t direct anyone.
Another shift that has occurred since the last time we reported on analytics salaries is the convergence of online and offline data analysis. Over the past few years, more and more companies are moving toward jobs that are not pure digital analytics roles. They are hiring people that can combine digital data with offline data sources to create richer marketing and customer insights. This is a trend that seems likely to continue, so we broadened the category to “Digital / Cross-Channel Analytics Professionals.”
Finally, our business has expanded. When we last reported analytics salaries our business was still 90% focused on digital analytics. Today we cover a much broader range of analytics positions, including: data scientists, analytics developers, statistical analysts / predictive modelers, etc.
Over the last few years we have developed a deep enough database of Predictive Modelers / Statisticians across the nation to report on their salaries, as well.
One note about “Years of experience:” For both Digital / Cross-Channel Analytics Professionals and Predictive Modelers / Statisticians years of experience applies to the number of years of experience IN THAT DISCIPLINE. For example, many of the digital / cross channel analysts used in this survey may have come up the marketing or web development ranks. This survey does not count those years of experience… only those that were spent in full-time analytics roles.
Finally, I want to thank Lee Feinberg of DecisionViz for putting together the Tableau Workbook. Lee is one of the top data visualization gurus around and we really appreciate his help.
Click Image to View Interactive Tableau Workbook
Click Image to View Interactive Tableau Workbook
12 thoughts on “Follow the Money: The IQ Workforce 2014 Analytics Salary Guide”
RT @corryprohens: Latest Blog Post – Follow the Money: The IQ Workforce 2014 Analytics Salary Guide: http://t.co/rB8NZbQzYb
Follow the Money: The IQ Workforce 2014 Analytics Salary Guide http://t.co/lBHMFKdlp0 #2014 via @corryprohens
2014 #Analytics Salary Guide #measure http://t.co/40uTbtsgPz via @corryprohens
@JoseAnalytics: A new report on #analytics salaries to start the year. Thoughts? #measure http://t.co/DSqISeiWYp
Do you realize your color-coded maps use nearly the EXACT colors used on blue-green plates in the Ishihara color-deficiency test? These maps are “perfect” for discriminating against people with color vision deficiency in the green part of the spectrum (including red-green and blue-green “color blindness”). Seriously, you just published diagrams that about 8% of your male audience cannot successfully process. This is something that definitely should not happen in a a profession where it is critical to communicate technical information to ALL of your audience, not just a random selection of most of them.
To make my point even more bluntly, CEOs are mostly white Caucasians, a population with higher color deficiency rates than many others. About 8% of them or more cannot even differentiate your diagrams. Nothing makes a potential client (or any human) “happier” than being treated like their disability doesn’t matter, especially when it is easy to avoid such treatment by simply using different colors. Please stop excluding color-deficient people from your analytics results. Nearly every person with color-deficient vision *can* see colors, they just have trouble with pale or dark colors that require green or red differentiation. Please respect that.
No, I did not realize that. Thank you for pointing it out.
One point of clarity on my post: even I can tell the difference between adjacent dots on your color legend and maps. However, matching the colors on the maps to the legend is imprecise because the colors look quite similar. Even a person with normal color vision might have trouble with a color legend if it uses similar colors and the legend lies too far from the color data.
As a father to a child with a visual perceptual disorder I understand how frustrating this sort of thing can be. Again, I appreciate you bringing it to our attention.
You are absolutely correct that this touches on the yellow-blue color-blindness scale; to my knowledge, this is “safe” for red-green. I actually prefer to work with blue-orange which is overall safe, and that’s what I’ve used on the other view.
Using these colors was a conscious choice more for aesthetics (I tried many combinations with this data set) — please let me explain. On the hierarchy of how we process visual information, size is easier to interpret than color. In this visual, size is the primary navigation for readers. Also, when working with the view interactively, you can click one or more circles in the legend to highlight them on the map. I hope that you are still able to gain some valuable insights from this important set of research.
Thank you, Lee. It sounds like you are aware of this issue (hence, the esteem your colleague bestows in the article). Shame on me, then. I apologize for my misinterpretation.
I am honestly surprised to hear you say there is no green component to the scale (yellow-blue). Given the trouble I was having with the colors, I must have subconsciously translated this to mean the blue had a slight green component that I was having trouble seeing. Often, that subconscious interpretation of my perception is correct, but only due to context, not pure color interpretation. For all I know, the difficulty I had is due to my monitor color (ASUS ProArt PA246Q set to 6500K, 2.2 gamma, saturation and hue both 50%, etc.) and ambient lighting (yellowish 60 watt bulbs, egg shell walls and brown furniture). No matter what, I instinctually allowed my color vision adaptations to lead me to assume that there was green component. Since a green component is not present, then It seems the colors used on this map are simply too close for me to easily discern them. I even tried a pair of “3-D” red/blue eyeglasses using alternate eyes closed, but the colors still were too hard to discern.
I have been frustrated with color use in presentations over the years. Despite awareness of this problem – the majority of coworkers and superiors have often done nothing to avoid ongoing problems. Many thought it was adequate to solve the problem by ad hoc instances of telling me which colored line or data point was which (and only after I asked). I was disappointed that other men (and once a woman) would speak up after I did and say they also had color vision problems, but they would have otherwise said nothing otherwise and preferred ignorance of the data over drawing attention to themselves. Thus, I feel like this is a solitary battle I fight instead of an awareness level I can simply add to.
Thank you for your considerations in designing with a yellow-blue spectrum. I’m sorry for taking my own disability and assuming the problem was your fault, and not my own (incorrect) interpretation of colors.
Pretty neat – analytics on analytics! Considering your national audience, the map of pay rates in different cities might be more useful if it also reflected the local cost of living – in other words normalizing pay as NYC and SF Bay look like a great place to go until you consider all the inflated costs of housing, etc…it would be pretty straightforward to use
Here is an example: How Crains Got Local Tech Worker Pay Wrong”
Is this data from your candidates only or do you have additional sources of data from which you have based your findings?