Hacker News Reader

Comment by jcadam | Hacker News Reader

Years of experience tops out at "8+"? An engineer that started working at 22 is barely 30 by the time they reach that experience level. Is 8 years really the point where additional experience ceases to have a meaningful effect on salary? Or do these three cities (SF/NYC/Seattle) just do a pretty good job of expelling software engineers after they attain their first gray hair?

View on HN

Easiest explanation is triplebyte doesn't have enough data past this. Especially when sliced by other dimensions.

At the big tech companies, you normally reach the "terminal level" (i.e. there's no more up-or-out pressure) within ~5 years. It's entirely possible to reach that level and then keep working another 30 years without a promotion. And even if you do strive to continue getting promoted, each level is significantly more difficult and could take another 5+ years to reach.

At 8+ you leverage your professional network and don't rely on websites like triplebyte

It may be the case that for a variety of reasons, salaries don't creep up that much past 8 years as it's just more difficult to either signal or create value then.

Despite the fact that a polygot dev. with tons of experience is always good to have around, and you need some weighing in on big decisions ... most dev is just dev - and a competent mid-tier developer with good habits can do the job in most cases.

In most careers, you top out around 8 years of experience for individual contributor. After that, you add value through scale of managing direct reports.

Yes, exactly. Diminishing marginal returns after 8 years.

That said, I really wonder about management - a talented senior dev who can write clean code, is polygot and knows their way around everything, gives good estimates, communicates well ...

I think that might worth more than anything really.

Often tech is more like a sports team where individual contributors matter more than those making other decisions, i.e. coaches.

(That said, there could be a serious market inefficiency here: a spectacular coach may be worth more than great players, it's just really difficult to measure!)

It's the latter.

I don't think that's really the case, but I do think it's common for good software engineers to move into other job titles (whether it's "founder" or "management" or "architect" or "software consultant" or a deep specialization in a particular subfield of software like security, deep learning, data science, data mining, compilers, databases, graphics, etc.) where they can avoid the salary cap. There's an adverse selection effect on those that don't; employers are likely to ask to themselves "Why are you still making $150K/year as an employee rather than $500K/year as a consultant, $2M/year as a deep learning specialist, or have retired as a founder? Did you never develop any sort of deep expertise worth paying you for in your 8 years as a software engineer?"

People going into software should plan accordingly. You get great wages to start, but you need to save some of those for a potentially risky mid-career transition where you specialize in a particular subfield.

That is only true in the HN bubble of SV. No one would bat an eye at someone only making $150K in almost any major city in the US outside of the west coast and NYC.

It's not the latter. Additional years don't matter as much past 10 or so... you're kind of doing the same types of things over and over as history repeats itself.

>you're kind of doing the same types of things over and over

Well this would be the root of the problem right here, not "experience doesn't matter after 10 years." That's nonsense. Experienced people with in-depth, cross-functional knowledge obtained from several, 5-8yr "sub careers" in various specialties can be worth their weight in gold.

I'd argue that when it comes to writing code, technology changes fast enough that after a decade, you've hit a pretty clear plateau. You've got the fundamentals down, and you're a master of your current language.

There's a lot of managerial things you continue to learn, but that requires you to be in a position with enough political capital to actually stop history from repeating itself. It's only useful knowing that Project X will fail, if you can convince people to instead do the more successful X++. I'd expect job title to be a much better indicator here: having 20 years of experience doesn't mean you have the aptitude or inclination to succeed at office politics.

That's assuming the company you have the experience from values titles. Ive worked at some that dont. A junior dev has the same title as the most senior dev: Developer. You cant tell from the title alone that one is making 80k and the other is making 500k+. Ive seen other companies go to the other extreme where the titles get ridiculous. Titles aren't a great indicator, in my opinion.

Fair - I meant title as a proxy for the work that's actually being done. My point was mostly that raw experience is only meaningful up to a certain point.

To be a bit longer-winded:

If you want to get paid more, you have to demonstrate value. Even if you only ever write the same CRUD apps, you'll definitely improve for the first few years, and it makes sense to pay more for a developer who isn't still busy making their first mistakes.

Past a certain point, though, raw experience isn't automatically a value add. If you're still just writing the same CRUD apps, you cap out a certain point. Now you need to demonstrate that you can contribute to architectural discussions, get involved in system design, or help with project selection/management. If you think you deserve the big bucks for that level of highly experienced work, that's totally fair, but you need some way to demonstrate it - whether this be tiles held, references, or having literally written the book on the topic. You can't just lean on "years of experience" anymore, because it's no longer well-correlated to value-added.

Reply on news.ycombinator.com