Are Only 20% of Devs Happy? | Stack Overflow’s Erin Yepis

0:00First of all, this one, I'm not surprised. Like I was surprised by only 20 % of developers are happy. I'm not surprised that one of the biggest complaints is technical debt. Like I think that's been since the beginning of time. I think that's what the data shows like on the linear B side. Not a lot of companies are, let's say, properly advocating or allocating enough space to reduce technical debt. So all of that makes sense to me.

0:35Hey, everyone. I'm your host, Ben Lloyd Pearson, Director of Developer Experience here at Linear B. And today I'm excited to be joined by Erin Yappes, Senior Analyst of Market Research and Insights at Stack Overflow. So Erin, thank you for joining us today. Thank you for having me. And I wanted to make a quick mention. I love the name of this podcast and I have listened to a couple of episodes. and I like the content of this podcast too. So really appreciate being on. Wonderful. I always love getting a compliment at the start of my day. So thank you for that. So I wanted to bring you on because Stack Overflow has this developer survey that comes out every year and we'll have a link to that in our show notes for our audience.

1:20But I heard you actually on another podcast discussing this survey and just felt like we had to get you on because it's such a relevant topic to our audience. So, you know, in particular, like we think about developer productivity, developer experience, like those really come to mind a lot. And I think there were some like pretty fascinating and maybe even alarming takeaways about developer happiness and job satisfaction. I was really looking at like the professional developers section of the survey because, you know, a lot of our audience are people who do software development professionally.

1:54And the thing that really stood out to me was one of the findings that said that only about one in five professional developers are currently happy with their job. Can you break that down for us? Like, what does this tell you and what does that number mean? Definitely. And I'm happy to talk about this because this is a new question. this year for the 2024 survey. You know, part of my job looking at the survey responses is to figure out the why behind the numbers. And so this new question gave us an opportunity to look at our community, which we also, our community is mostly professional developers as well, and then ask them how they feel about work.

2:37We have, we've run surveys in the past where we ask directly developers how their jobs are going. We asked them some questions about what you look for when you're searching for a new job. What are some of the reasons why you might leave your current job? But this would be the first time we're asking just blanket statement, umbrella statement. Are you happy with your current job? So I definitely thought it was really cool. And so also we know in the background of a lot of the research we've been doing too, that there's been disruptions in the industry. so it was top of mind as far as like the satisfaction goes, job satisfaction.

3:17For this question in particular, we used a NPS style scale for, which means users could select from 0 to 10, happy, unhappy, 10 being happy. The NPS style I thought was a good option for this question because unlike the Likert scale, which is five options, the NPS relies a little bit more on the normal distribution, assuming that a lot of people are going to be picking something closer to the middle that way if you are categorized as a promoter or happy happy developer in this scenario you selected nine or ten and i also like this too because so for anyone that was maybe a little wishy-washy on the fence um it's like saying if you're not a hell yes you're no i thought it was Yeah, you know, sometimes the surveys with a qualitative approach, you have to be a little bit tricky in order to get the real scoop.

4:18Yeah, interesting. It's surprising to me seeing that, you know, I think that same number was about almost a third of developers were unhappy with their work. So, like, there definitely seems like there's this large volume that's sort of in the middle. You know, I think the word you used in the survey to describe it was complacent with their jobs. but you know it's just surprising to see that 80 percent of developers basically are really not engaged with their work and to me that that goes against my perception of the field you know because i know it is a tough uh profession to get into but you know at the same time it's it's a it's a pretty good job to have you know being a software developer so you would you would like i personally expect it to be very different you know not necessarily like developers all being super excited about their job, but, you know, seeing only one in five, like, really engaged.

5:08Another interesting aspect of this was that, you know, their managers actually weren't even much happier, because I think that number was a little bit closer to, like, one in four managers are happy with their work. You know, it seems like they're similar, but still a little different. Do you think there's anything to that? Yes, I do. So I think that, and as far as, like, just to go back to your comment about the expectations gallup also put out a similar poll not just for software developers but a global poll on engagement at work and it was also pretty low so i definitely the number seems believable to me it is a little rough i've seen a couple reddit threads where People go into the depths of how they really feel about work and specifically about this stat from the developer survey.

6:04So I was able to, I can tell you that I looked at, I wanted to look at the numbers in relation to salary as a hypothesis might be that salary had something to do with job satisfaction. So I was able to see that expectedly salary is correlated with satisfaction across globally, not just in the United States. But this relationship was more or less strong in certain countries. When I looked at the top 10 countries that responded to the developer survey, I saw that most happy developers are definitely above the 50th percentile for salary. and developers in the Netherlands in particular, regardless of whatever percentile salary they were in, their job satisfaction scores stayed consistent.

6:57So they were less swayed by higher or lower salaries. However, those developers that were in Brazil or France were definitely more swayed by higher salaries. I also looked at, as far as countries and the effect of workplace setting. Interestingly enough, Germany, India, and the UK, we saw their job satisfaction scores were more affected if they were in a lower salary percentile and also in-person work setting. And then just going back to also your comment about the managers, the people managers, engineering managers, as far as the role goes, had some of the highest job satisfaction scores of all of the roles that we asked about in the survey.

7:49So regardless of their pay, their job satisfaction scores didn't really move that much. The same goes for non-people managers, data engineers, desktop developers, and embedded developers. Their job satisfaction scores kind of stayed pretty consistent regardless of their pay. Interesting. There's almost like an anthropological study to be done here on the differences of all these cultures and how they view happiness within the microcosm of the software developer universe. Well, yeah. I mean, you've heard about, I think Germany takes their time off very seriously. I would expect the Netherlands does too.

8:34Also, the Netherlands, I'm sure when you go into the office, it's like a nice place. Maybe there's water. nearby yeah and you you wrote your bike to work you know yeah it was something there yeah i definitely you know my sense of of working with a lot of companies in the netherlands is that they definitely are more like office centric like everyone bikes into work to the same place they you know the the there's a lot of benefits that that dutch citizens get you know so i think it does kind of make a lot of sense and you mentioned that there there aren't previous this is the first time you've asked this question, so there aren't previous years to compare this to, but I wonder if there are any other data points that you've measured in the past that we can maybe see if there is some sort of trend happening.

9:18Like, is this a post-pandemic thing? Is it something that's always been consistent? Like, do you have any sort of indication like that? Definitely. So the closest thing I have, like I mentioned, were the job surveys that we've done for Stack Overflow, we have specifically the ones that we did in 2022 and 2023. We ask developers to tell us, you know, of all of these facets of work, what are the most important ones to you? In both 2022 and 2023, flexibility ranked higher. It ranked absolutely the highest in 2023 over salary. So salary is up there. Salary has always been important. The flexibility was up there.

10:03And then rounding out the top three learning opportunities always comes up pretty high, too. So I definitely expected salary to be as anyone would. Salary makes a difference as far as if you're happy with your work. But because of COVID, I thought the workplace setting and flexibility, being able to pick up your kids, taking time off if you're sick, that that was more important, too. So we can see in the developer survey that there are, it's a slight shift away from fully remote positions to hybrid positions, which we would expect not a big jump in in-person work settings for developers. But we do know that there is a big difference year over year for salary globally.

10:51And I think that is our job survey results combined with what we see in the salary trends is what makes me think that's the biggest thing we have to compare. Yeah, interesting. And I wonder if COVID pandemic may have made that type of flexibility more commonplace. So now it's just assumed that it's a part of new jobs a developer might take. So to them, it may still be very important, but they just don't think about it versus salary. particularly given that we are in a bit of an inflationary environment. So it almost does mimic some macro trends in a way. Is there any other insights that you think that we're getting from this into why developer satisfaction might be so low?

11:40I think going back to the roles, the salary percentiles, and the country trends we saw among the happy and unhappy developers, we know that or at least I was able to see that amongst engineering managers like I said who generally had a higher job satisfaction than any other role and it wasn't swayed as much by salary or country that they overwhelmingly indicate that compared to other job roles that driving strategy is a factor in what they like about their jobs which again yeah separately with the job satisfaction question we asked what factors contribute to your being satisfied at work so overwhelmingly we know that most developers indicated that improving quality code was what they most found enjoyable about their work but yeah so driving strategy for those engineering managers and then they also indicated overwhelmingly that being a power user was less important to them so i think that and And then so to go along that with those roles, embedded developers, they highly rate improving code quality higher than the average developer.

12:59I think the desktop developers also, they highly rate contributing to open source code. I think what that shows is that certain people get into the game for a reason. They find their niche. and then if they're able to really play around in that area, that thing that makes them happy, improving quality code, driving strategy, oh, embedded developers, also really liked working with high quality hardware. Makes sense, right? So we actually have numbers to go with that with something that intuitively makes sense, but there are some of these questions that we don't get a chance to directly ask, especially to an audience this big, and now we have data to back that up.

13:41Yeah, I like the insight about engineering leaders or the managers wanting to be more focused on driving strategy. And I think it actually makes a lot of sense why that would be such a big focus area right now. Because, I mean, you think about the last year, you know, again, coming out of the pandemic, we have all these companies that overhired that, you know, we're building up business during low interest rates, low interest rate environment. And now we're hearing this from company after company where engineering teams are either doing the same with less or being expected to do more with the same rather than increasing headcount to hit their goals.

14:19And I think that makes it really difficult to drive strategy when you're in that kind of environment because you're more focused on just keeping things operating rather than being strategic. You know, getting a little more into these factors that contribute to job satisfaction, you broke down a lot of things by like things that contribute to happy points, complacent points, unhappy points. So what are the major takeaways that you've seen from that section of the analysis? I think the improving quality code overall just came up as the top factor for job satisfaction. and I definitely, that makes sense.

15:02And we love to see that. And again, so seeing that driving strategy actually made no difference to a lot of the developer roles makes sense. So there are people that kind of want to focus on improving quality code. I don't want to drive strategy. I want to do this thing. I like to see that. I like to see that difference amongst the different groups. What was the other thing too? I saw, because again, the Netherlands just kept coming up. The Netherlands. So regardless of the salary or your job satisfaction score, again, very consistent. Netherlands developers highly rate designing architecture and environments.

15:45Again, it makes sense. I think there's definitely a bigger story here. It has something to do with why you decided that this is what you're going to do for a job. So there are so many different things to go dig into about that. I like the idea of using this data to explore reasons for the why, rather than focusing on the overall number, like doom and gloom, this is terrible. Again, kind of going back to some of those Reddit threads that I saw about these results, the one in five. There are some types of people that are like, they see that number and it's a confirmation bias. Like, of course, everything is terrible.

16:26Of course, it's like this. But that's not the reason to ask about this and to quantify it. The reason to ask and quantify it is to find out why and to do something with that information. Yeah, exactly. And I think the improving code quality, I mean, that comes up with like in our world a lot as well. And I think it's part of this challenge that virtually every single engineering team has. They want to do things like reduce tech debt, improve developer experience, reduce toil and maintenance and all of that. But the problem is there's the rest of the business that has all these priorities that are constantly pressuring the engineering organization to deliver on.

17:04It's like building new features, enhancing things, building stuff for new enterprise, large clients or whatever. And a lot of engineering leaders, I think, actually struggle to push back against the rest of the business to say, we can dedicate a certain percentage of our resources to solving, to building the stuff that the business needs. But we still need to have some portion of that dedicated to improving the things that make our developers' lives better. And I think because that conversation is so difficult for engineering leaders to have sometimes that a lot of times things like code quality get deprioritized within the organization.

17:43So were there any surprises or things that you weren't expecting about the stuff that contributes to happiness? So I'll go back to the salary correlation. While I know salary is important from other research that we've done, again, my mind was that in the space that developers, at least in 2023, valued flexibility more than salary. So I know that, yeah, inflation, economic headwinds, and then seeing the numbers that people were reporting for at least a 10K, if not more, USD decrease in annual salaries for all roles, most roles. I was thinking at first that maybe had more to do with developers because they prioritize that flexibility, that remote setting or hybrid setting that they were choosing to stay in those roles.

18:48But I think we saw with the job satisfaction and then just like, yeah, running that analysis on workplace setting and salary, the nope salary is much more correlated than workplace setting. So that was surprising to me. But again, after you think about it, it does make sense. Yeah. I want to take a moment just to talk about some of the frustrations, the common frustrations that you brought up as well. And I think we touched on this a little bit already, but the top frustration that was surfaced in this is technical debt with over 63 % of respondents saying that that contributed to their frustration.

19:27So what are your findings around technical debt? Is there anything deeper into that that you can share insights on? uh yeah this is a good one i i think my main finding and i mean to be very transparent is that because it was so highly ranked amongst all the choices for frustrations that my main finding is that tech debt is just too broad of a term to be using for that type of question so that's fine You live and you learn. I definitely expected there was going to be more parity amongst those frustrations because work is frustrating sometimes, especially if your work is very complex. But tech debt, I think, is too much of a buzzword at this point to actually mean anything specific.

20:20So I think I, for the next developer survey, need to tease that out a little bit more to find out. Interestingly enough, we had a really great episode on the Stack Overflow podcast about tech debt mindset that came out after the results for the annual developer survey released. So I got to listen to that and it started making more sense. The phrase tech debt, technical debt, is its own microcosm. And it has to be, you have to be more specific. I know that now. So I think, yeah, I think to make that question better, the idea is going to be to reformat it to ask about causes of tech debt, like in the terms of frustrations at work instead of just blanket tech debt.

21:12Yeah. So you're thinking like almost get more tactical, like is it outdated dependencies or lack of tests or something like that that's contributing to it? Yeah. honestly i'm gonna have to do more research of what people consider that might be its own little mini survey like so complicated i guess yeah and and i also loved you know i feel like this always shows up but i think like some of the top ranked stuff also included like frustration with like build and deployment systems which you know i'm still here wondering why in 2024 haven't has no No one figured out CICD. Like, we're still struggling with those things.

21:53I feel like I'm still figuring it out, too. So, yeah. Well, thank you for breaking all of that down for us. So let's start talking a little bit about, like, what organizations can do with this information. So if you have an engineering manager, director of engineering, or any sort of engineering leader reading this survey, like, what would you hope that they could take away from this to help improve developer satisfaction? I would start with read the developer survey, go visit our developer survey site. That's a great place to start, knowing that this is the largest developer survey in the world and it's a global audience.

22:30So you're getting a lot of perspectives. But besides that, I think we saw salary is a real problem that is just correlated to that unhappiness score. I think that some of these things, again, like I mentioned before, seem intuitive, maybe a little obvious, but there's numbers to back it up now. Besides that, I think with the, as I mentioned before, with those factors for job satisfaction, going back to there is a connection between your role and the things that you do at work. So if you're a people manager, an engineering manager, I think checking in on what it is that you like to do is always a good idea, regardless of the numbers in this survey.

23:22But what we're seeing is with certain roles, there is a relationship between your role and the things you like doing. And that thing you like doing is probably in the job description of that role to begin with. Awesome. Awesome. And then to flip that, what about like an individual contributor? Like, is there something from this that you think they could surface to their leadership? One of the enlightening things that individual contributors responded with in the survey and could bring up to their managers was that most developers do not feel threatened by AI. that we had 70 % of professional developers indicated they didn't believe AI was a threat to their job.

24:06So given that there is maybe some uncertainty with like job security or uncertainty with how roles are going to be affected in tech orgs in any industry with AI, that if it's something you don't feel threatened about, Then talking to your managers about how can we use this? How can I use this tool? How can I get ahead of what these tools can bring, how they can improve? I want to be using them. If it's not a threat, then I think having direct conversations about it makes sense to me. Awesome. I want to maybe pick your brain on what's coming next year as much as we can. So, you know, I think ways that I would love to learn more about some of these insights that you gained is, you know, one of my favorite parts of the developer survey is always the most loved versus like most wanted tools.

25:04Like I would love to see like correlations between that and like people who are happy with their job. You know, are they using the tools that everyone wants? You know, that's my idea. But like, do you have any, you know, you mentioned breaking out tech debt a little more. Do you have any other plans for how you might be able to expand on this professional developer portion of the survey for next year? Yeah, and I do think that's a good idea. The one problem with that is there are so many technologies that we asked about. And so there's a lot of noise there, I guess. But I do love that. I think there is some way you've got my wheels spinning now.

25:44So I'm going to, I will take that back. And obviously that's our favorite part of the developer survey too, seeing the results of who's using what, what do they want to use. So I guess for next year, we go through this process where we source the technology selections that will show up on the dev survey. and so one of the differences one of the new things we did this year we we always go to it's a combination of looking at some our internal data sources so we have tags on stackoverflow.com that we can see trends in we have external data sources such as github and google search to see like what are some of the technologies that are out there that people are that are popular in those spaces.

26:36And then we also ask our meta community for feedback. Hey, these are the technologies we're thinking about asking our users this year. What do you guys think? And they always give great feedback. They love to correct spelling. First of all, that's always important. And but this year, some of the feedback that they gave was, hey, you have a lot of you're missing some of these embedded technologies. Embedded developers accounted for 3 % of the roles that we asked for this year, which is not a small number when you think about all of the roles that respond to the developer survey. So we, I went ahead and created a whole new technology subsection just for embedded tech.

27:21So I think there's possibility we could see more stuff like that. I think it just makes, it depends on some of the feedback, what we're seeing as far as like those data sources go, like sometimes we have to take some of those technologies out of the list too like hey this has been slowly dropping out of people's usage and they don't love it they're not using it they don't want to use it i hate to say it but you know my one of my first languages my most loved language is falling down that path as well i hope that turns around but it's not up to me so for the professional developer series i think for questions i think probably not too many changes just because we like to see that like history that that linear time series of how these answers are are progressing over the years but we are definitely attuned to like what is going on with ai what is going on with changes in the workplace and how are those being reflected because we want to specifically ask those either new, newly professional or seasoned professionals, how is work going for you?

28:39What does work look like? So we'll see. I think it just, it remains to be determined what else is in store for us, like what new progressive technological expansion is knocking on the door of 2025. Yeah, I think in particular, there's a lot to be learned about the impact of AI on different types of developers. Like, I think we, you know, with a lot of the tools that are coming out, it's like we've sort of, like as a whole, just view it collectively as like, well, all developers are now using co-pilots, so they must all have a similar experience with it. But I think we may be learning the reality is that like a senior developer has a very different experience with a tool like that than a junior developer.

29:20And they both could potentially see massive improvements from it, But you may have to deploy it in a very different way, depending on the person. So, yeah, I think, yeah, any, you know, everyone's talking about AI for a good reason. But I think it is very relevant that, like, we need to understand how it's going to change not just the whole software development industry, but each individual role within the organization. Yeah. And also to go along with that, you know, we have we ask developers now what part of their their development process do they use if they use AI tools that they use it in.

29:56So we have tons of options, but similarly to this job frustration and tech dub being overwhelmingly the top one, developers indicate writing code as their like top, overwhelmingly top way to use AI tools. and I do think that maybe needs to be broken out a little bit too because it's just a little too broad. And exactly going back to what you're saying with junior and senior developers, they're writing code, but very different. This has all been wonderful, Aaron, and thank you so much for joining me today. And again, we'll have the survey link in the show notes for all of our listeners. And there's tons of information that we didn't even get to touch on today that I think, you know, plenty for people to learn from within there.

30:45So I definitely encourage our audience to go check it out for yourself. So Erin, if people want to follow you to learn more about what you're working on, where should they go? Yes, thank you for asking. I would say, please check out the Stack Overflow blog because I post my analyses there. We are currently working on a survey for our community members about knowledge in online communities. So that analysis should be coming up soon. And depending on when this podcast is released, the survey may still be open if anyone wants to contribute. That would be cool. Wonderful. Stick around. After the break, Linear B co-founder Dan Lyons will be joining me to discuss today's episode.