This report is based on a survey of 65,000 software developers from 186 countries around the world. This is the number of responses we consider “qualified” for analytical purposes based on time spent on the full, completed survey; another approximately 400 responses were submitted but not included in the analysis because respondents spent less than three minutes on the survey.
Other (country not listed)
The survey was fielded from February 5 to February 28.
The median time spent on the survey for qualified responses was 16.6 minutes, down from 23.3 minutes last year.
Respondents were recruited primarily through channels owned by Stack Overflow. The top sources of respondents were onsite messaging, blog posts, email lists, banner ads, and social media posts. Since respondents were recruited in this way, highly engaged users on Stack Overflow were more likely to notice the prompts to take the survey over the duration of collection promotion.
As an incentive, respondents who finished the survey could opt in to a “Census” badge if they completed the survey.
Due to United States transport/export sanctions, our survey was unfortunately unaccessible to prospective respondents in Crimea, Cuba, Iran, North Korea, and Syria, due to the traffic being blocked by our third party survey software. While some respondents used VPNs to get around the block, the limitation should be kept in mind when interpreting survey results.
In years past, our analysis of professional developers was based on site activity on Stack Overflow. This year, we utilized answers regarding employment to deduce whether or not a respondent qualifies as a professional developer and built our analyses based on this qualification.
We asked respondents about their salary. First, we asked what currency each respondent typically used. Then we asked that respondent what their salary was in that currency and whether that salary was weekly, monthly, or yearly.
We converted salaries from user currencies to USD using the exchange rate on 2020-02-19, and also converted to annual salaries assuming 12 working months and 50 working weeks.
This question, like most on the survey, was optional. There were 34,279 respondents who gave us salary data.
The top approximately 2% of salaries inside and outside of the US were trimmed and replaced with threshold values. The threshold values for inside and outside the US were different.
Many questions were only shown to respondents based on their previous answers. For example, questions about jobs and work were only shown to those who said they were working in a job.
The questions were organized into several blocks of questions, which were randomized in order. Also, the answers to most questions were randomized in order.
To identify which technologies to include on the survey this year, we looked at both the most popular and fastest growing tags on Stack Overflow (in terms of questions posted). We compared these to the technologies we included last year and looked at how many people chose each option. We synthesized all this together to curate a collection of technologies to include.