10 years of Stack Overflow Surveys

Ismael
3 min readAug 25, 2020

--

A brief look into the evolution of job satisfaction, programming languages and developers’ salaries

Life would be cumbersome for developers without Stack Overflow. It has become the de facto source of answers for tech-related questions (after any official documentation, right? Sure, why not).

Since 2011 Stack Overflow has been conducting surveys related to a broad variety of topics, ranging from work remote status, world location, and preferred technologies, trough more lightened questions like the way the word “GIF” is pronounced or how often developers get “into the zone” when coding (what would we do without those extra pieces of knowledge?!).

This lets us know more about the community. What matters most to them, their level of satisfaction, or what they are excited to learn next. Stack Overflow shares its results analysis as well as the original survey datasets. You can find them both in detail here.

These surveys are conducted every year between mids January and February. So for this year, results haven’t been influenced by the COVID-19 pandemic.

The following summarizes the evolution of three categories, mainly top programming languages, job satisfaction, and developers’ salaries.

So, first things first.

JavaScript never gets old

During the last 10 years, JavaScript has remained as one of the top technologies used by developers. It’s the language of the web after all. It is also the place where beautiful visualizations come to life with D3 and p5.

Programming and Scripting Languages over the years. Percentage of respondent users.

At the top of the list are also HTML and CSS, complementing web technologies, SQL query language for databases, and Python as a multi-purpose programming language.

(Can’t get No) Satisfaction

Well, developers can. Over the years developers’ job satisfaction has kept high with a proportion of high satisfaction above 60% of professional developers that participated in the surveys.

Respondents’ Job Satisfaction. Full size here

Bellow is a Theil’s U plot, an asymmetrical measure of categorical association. Analyzing 2019’s survey, the possibility of working remotely (though by now is imperative) slightly stands out among the reasons that could explain developers’ satisfaction together with the influence of the main branch (developer by profession or someone who writes code as part of their work).

Theil’s U association plot related to 2019 categorical questions.

It’s also interesting to see that the education level or the undergrad major seems to have little if any relation with job satisfaction. So whatever your background is you have pretty much the same chances to have a great time at work.

Developers’ Compensation

Analyzing the year annual compensation density plot below we could see that the shape of the curves per year remains similar, with a tendency of normalization in lower range salaries.

Kernel Density Estimation plot related to annual compensation per developer type.

Depending on the developer type, curves get skewed to lower salaries (Mobile Developers) or better approximating to normally distributed (DevOps).

Engineer Managers keep reporting the highest annual compensations, with median salaries around 90k USD dollars per year. For the rest of professional developers, salaries range around 55k per year.

Median compensation per developer type for the last 3 years.

So, what to expect in the future?

COVID-19 has changed the game for everyone, that’s a fact. A change in the features analyzed is more than expected.

We’ll have to keep track of the trends in the industry and the evolution of the current events to take action accordingly.

One thing is for sure, the community will keep being awesome, so as you.

That should be all for now. I’ve published a more technical GitHub repo (vastly turning to Stack Overflow for answers during the analysis), so if you want to know more you can give it a look.

Have a great day you.

--

--