Given it's a DigitalOcean survey and the data was collected from customers and via social media sites, it could always be sampling bias. The people who answer s

Given it's a DigitalOcean survey and the data was collected from customers and via social media sites, it could always be sampling bias. The people who answer surveys are generally a much smaller group than the population at large, and usually more on the 'enthusiast' side of things. Add how they got the data from social media sites like Reddit (according to their blog post), and well, it's quite possible the people they surveyed are simply the minority that go on Reddit, read Hacker News and keep up to date with tech news rather than the majority of the developer/software engineer population. I mean, the majority of this industry doesn't consist of convention goers, open source hobbyists and hackers, it consists of people for whom software engineering is a day job in the same way as retail/finance/sales is for others.