Tech salaries continue to increase as firms search for talent
computerweekly.com
computerweekly.com
Right now my interest is spread thin and I don't mind doing any of them but I would rather do something that pays.
It's still a lot of money, but it's not as much as you think it is, since it's not just pure salary. Even if it was, odds are it would still be taxed significantly, because most high-paying SWE jobs are in California or New York. California has 7x SWEs as Washington, which I would consider a better state in terms of how much overall compensation you can keep after taxes. It's better, but there's just less jobs in comparison.
I don't quite get why there is this misconception that, at the end of the year, bonuses are taxed differently from regular income. Withholding by the company is what may change, but when I receive my W2 there is no special "bonus" section, the bonus is combined with other earned income and listed in box 1 of my and other people's W-2.
At a big company, I would guess, it would be the effect leetcode-based hiring. It's easy to build a solid team making $300k-$400k in total compensation where each know how to easily move knights on infinite chess boards and find substrings with largest variance but won't necessarily deliver good software or know how to work together well.
If you had to choose one engineer among 100 candidates using whatever approach you want but within a time limit of 10 hours per candidate, I am pretty confident you would accurately gauge the "quality" of candidates.
I think I could filter out the ones that are totally unqualified. Probably reliably end up with one in the top 50%; maybe even the top 25%. But one that's going to be sustainably effective, alongside the rest of the team members, through a variety of real-world situations? Nah. It's basically random. The same interviewee could turn out to be anywhere from "the best hire we've ever had" to "can scrape by but isn't very effective overall", and in many cases there would be no way to tell beforehand.
This is why some companies just give up and use Leetcode. If it's going to be mostly random anyway, they figure they might as well save time and effort. I wouldn't choose to go that route, but I get it.
Let's say you are asked to rate a boxer. You have their resume (who and when they fought) and you can "interview" (no sparring partner, even) them for 10 hours. If I am going to rate them on a scale 1-10, I could certainly rate them with an error of +-10% (on average), that is I could make an accurate prediction of their "worth" in the next year. It is not just the "interview" without sparring partners, it is the interview + resume. Only the interview and no resume, well, there have been a lot of promising boxers who went nowhere, and in fact it is much more challenging to evaluate juniors than seniors.
In my experience, I was able to rate SWE and ML types pretty accurately, conditioned on the fact that they were hired (and they hired some people I rejected and some I approved). I had them for one hour and I read the resume. I think the secret was that I asked good questions.
It's hard to separate a team's performance from the rest of the org, and it's hard to separate an individual's performance from the rest of their team. And even if you could- how do you measure success? Unlike boxing, there's no binary win/loss condition. KPIs? But what about an engineer who does a brilliant job quietly making sure a piece of software will be maintainable over the next decade? How do you measure that, even from inside the organization, much less outside of it? And if you could, how do you know that success in one org (with a particular codebase, stack, team culture, velocity, etc) translates to success in a different one?
And then because of all the above: even a resume doesn't tell you very much. It tells you they convinced other employers they were worth hiring/keeping, but those other employers were subject to the same imprecisions/measurement problems that you are. It's very possible they were all wrong about the quality of this person's contributions.
Software development is a deeply human activity that's impossible to reduce to simplified signifiers. It's my opinion that the only way to evaluate an engineer is to be another engineer working alongside them for an extended time. And even then- that doesn't necessarily translate to other scenarios they might work in.
Another commenter here said they do a 30-day trial period with new engineers, to see if they're a good fit. From a hiring perspective, this is the only process I can think of that would be decently accurate. Though obviously it would be pretty rough on the engineer to not know if they have job security for 30 days.
What I've seen happen is well-meaning people who don't understand software try to do the right thing as they've been taught, but that often means scrum. But then it turns into a very top-down approach and developers start getting treated like line-workers.
At the company I'm at now, "product" has wedged themselves between the scrum teams and the business side and they basically won't allow the two to meet. They wield power over the technical side as a result. I actually was in a meeting with business once and they asked the question "who should be involved". I responded product and technical so we can advise. The product immediately responded "product" and the lack of technical in that response was deafening.
So long story short is power politics or the developers aren't as good as being claimed (or a healthy mix of both). At the end of the day, these companies just have to not get it so wrong they go out of business. When the bar for that gets to a certain point, power politics takes over, and the ones who have business' ear are the ones who wield power.
In my industry, defense contractors are absolutely beyond desperate now. They spent decades moving their facilities to from very well populated suburbs of big cities out to low tax areas....in bumfuck nowhere and they more or less could force the boomer generation individuals to move there. Now, especially post COVID, they can't find anyone willing to sacrifice their quality of life anymore. To the point they have been opening up some jobs that can be as remote for security clearance jobs. Though I've heard of at least Boeing bait and switching individuals hard out of desperation.
It feels like some areas are still seeing demand for tech workers.
At my company, external hiring and internal transfers are all on hold, at least until next year.
Comp is going to continue falling, given it depends more on equity than salary. Projects aren’t really slowing down though, and we’re continuing to burn people out at an even faster rate.
I’m waiting for my upcoming equity grant. Going to quit, do nothing for 6 months, then look for a job.
Is it a good idea to quit just before the economy really tanks and most companies freeze hiring? Probably not.
I’m so burned out from this fucking, abusive company. I hope the government steps in, jails their executives, and breaks this company apart. 70% of this business wouldn’t survive if that happened, and these cold, soulless so-called “leaders” would be begging anyone for a job.
Amazon couldn’t pay anyone enough for the verbal, mental, and physical abusive it offers.
They have literally never changed, I could show you at least 5,000 emails and linkedin messages with these ranges.
If you dont worn at FAANG or a company in that range, then your salary has not even kept up with inflation in 20 years.
Talked to a recruiter yesterday who said “we have a generous package, we can pay up to $160k and give you 12 days of vacation, 15 days after 3 years”. In San Francisco no less.
FAANG salaries out outliers and we do a lot of harm by claiming that all techies are making $300k+
If you arent good enough for FAANG, texh is a dead end career. I know Assembly line workers who dropped out of high school and make as much as I do With my wasted 6 years in uni.