Leetcode interviews have spread throughout the industry and it has become increasingly hard to get a job offer now. Even startups are throwing LC mediums at candidates and expecting perfect solutions in 20min. Many companies like Facebook require 2 perfect LC solutions in 1 interview. It is very competitive and requires a ton of studying to reliably clear these interviews. Overall I think interviews are getting harder. Every company thinks they are google now and they’ve become obsessed nitpicking over useless toy algorithm problems. The increasing difficulty seems to imply an OVERSUPPLY of engineers. Too many engineers looking for too few jobs so the standards are getting jacked up. But then again TC for engineers has been growing well in excess of inflation for a while now. In the Boston area it’s been growing at about 7% YOY since 2010. This would imply an UNDERSUPPLY. It seems contradictory to me that both standards and compensation are increasing in lock step.
There’s oversupply but since every man and his dog needs bespoke CMS and ton of other crapware to move a button few pixels off for no reason - there’s a huge need for people to perform all this scaffolding
Women can be engineers too.....
This won’t be popular on this app, but just be ready for the bubble to burst. SWE is being commoditized and after the next recession, God knows when that is, there will be a glutinous supply of SWEs. Until then, ride it out and make that money. We’re at least a few years away from that point imho.
I think that can be said for any job during a recession. Sure if a recession hits SWEs will have a hard time finding jobs but I don't think it'll be hard compared to other non tech jobs
The supply of swes will be sticky? :D
Not really, there’s still a massive undersupply of software engineers.
Not according to our data 😏 If you’re interested in these trends, I suggest you to follow LinkedIn’s economic graph report. https://economicgraph.linkedin.com/
? What does your data show? Under vs over?
Depends on your city. In general undersupply you can look at any labor statistic to confirm. The problem is wages are growing and there is a glut of bad SWE's. A buisnesses only response given the cost is to institute guards to weed out the bad ones, however the leetcode method is not a good measure. Buisnesses will have to evolve to a better weed out method or find themselves always short on staffing.
What about hiring devs and firing the bad ones?
I agree 100% with that sentiment but it seems HR's and companies refuse to fire the bad ones no matter what. Which in my opinion is making the situation worse.
Great question! An above thread pointed out that there is an undersupply of good SWEs, which I think is true and the places I’ve worked want more. Another post pointed out that LC isn’t the best way to test SWEs, which is also true. But, most companies are using it to filter out the bad from the acceptable, and it’s hard for someone who is really bad t crank out code in a 2-4 LC style interviews that none of the sessions doesn’t catch as bad. So basically, the hiring process is trying to filter out the bad, not find the great ones. This also adds to the undersupply problem, because you need to hire more people since you don’t have a hiring process that hires the best. This does put us on a fragile situation though, if Google was better at hiring (and maybe firing) people, it would need less “good” SWEs since it would only keep “great” SWEs, so it’s demand would actually drop. It’s also not clear it would need to pay those “great” SWEs much more, although maybe it would because of Google could filter down to just the great ones, everyone would try really hard to poach from Google. The poaching thing also creates a weird incentive. Maybe you want to hire people that aren’t great to make it expensive for your competition to poach your talent because they need to interview them to filter the good from the great.
This may be unpopular but I only want to add that I would not consider a SWE bad off not being able to crank 2-4 LC's. It would be relative to level, company, actual work problems, and personality (yes none code related meatrics). This combined with interview performance would be a total score.
undersupply of Sr. level engineers. oversupply of entry level.
Leetxoding is a bad way of measuring sw eng quality
Major oversupply
Undersupply of good engineers. Money has drawn in a lot of shitty ones though. This is why the difficulty has gone up.
The problem is that leetcode is a terrible way to measure engineering quality...
If shitty engineers are entering the field just for the money by grinding LC, then why is the solution to up the difficulty by asking more and harder LC questions?