4 ms·
Insightful comment overall, but I disagree with your implication that the formula for pay raises is generally "current market value minus the cost of switching
by dunpeal 8y ago
Insightful comment overall, but I disagree with your implication that the formula for pay raises is generally "current market value minus the cost of switching jobs right now".
First of all, what "cost of switching jobs" are you referring to?
If I live in SF, "switching jobs" likely means getting off at a different BART station. Hell, I might be able to walk to my next job.
I think the well-known fact that pay raises fail to keep up with market wages in tech - isn't some brilliantly efficient calculation like "current market value minus the cost of switching jobs right now" as you suggest.
It's clearly not working very well, since job hopping in the valley is crazy high, and everyone knows that's done because of pay. To wit, the longer you work for the same company, the farther your market comp will drift away from you, and the only way to catch up with the runaway comp-train is to hop jobs.
I think it's simply the traditional American 2-3% raise model clashing with a booming tech reality where your market wage will likely increase much higher. Plus, the traditional model itself never worked very well for well-performing employees:
> It used to be that jumping ship meant landing a salary 10 percent to 20 percent higher than your previous one. While increases of that size aren’t as widespread as they used to be, switching jobs is still the most common path to the best pay raise.
> If you stay at the same organization, your annual increases may be restricted by your current base salary because companies have a narrow percentage range within which they can boost your pay.
https://www.investopedia.com/articles/personal-finance/090415/salary-secrets-what-considered-big-raise.asp https://www.investopedia.com/articles/personal-finance/09041...
Simply stated: employers were never very efficient at raising comp for good performers. Even decades ago, in industries nowhere near the growth rate of current tech, top performers were much better off job-hopping. Tech just made it even more lucrative.
- geofft 8y agoBy "cost of switching jobs" I mean that there's a negative cost in the long term to having a short job on your resume (especially if your previous job was also short), to not having an enthusiastic positive reference, to maybe getting a reputation as someone who doesn't want to stay at a job, etc. If I take a job paying, say, $100K, and someone is offering me $120K after six months, waiting until I've been at the job a year or two is technically costing me $10-30K over jumping right now but is probably worth more than that in the long term just to not have that six-month stint on my resume and to not burn bridges with my current employer. After a couple of years it becomes more defensible to leave, so there are mechanisms for giving you significantly more money after several years (promotions, promotions to manager track, refresher grants, etc.) but fewer mechanisms for giving you more money in the first year or two, and basically none for within the first year, because employers know that even if your market value is theoretically higher, you can't actually take advantage of that yet. I think that even in markets where job-hopping is high in general, any particular individual can still only job-hop so many times before it backfires on their ability to get a high salary.
- dunpeal 8y agoHopping after 6 months, or say less than a year, will definitely hurt your resume, even in the Bay. As a hiring manager, I'll typically assume your either got fired, or gave up too quickly. Both strong reasons not to hire you. Hopping after 2 years in the Bay is extremely common. So much that I'm not even sure it's considered "hopping" anymore. I'm not sure how long of a time horizon you mean when you suggest job-hops will eventually hurt you. In the Bay, at least, if you have say 5 2yr stints, what's likely to hurt you is... your age. A developer with 2 2yr startup stints is your typical Bay "senior developer". And a developer with 1 2yr startup stint is your typical "experienced" Bay developer, and generally one of the most in-demand categories of developers in the Bay. I'd go so far as to say that the 2 2yr stints will look better to a startup hiring manager than 7 years in an corporate environment like Oracle. The former is definitely in tune with SFBay startup-scene vibe, the latter looks like a corporate potato whose fit for startup life is questionable (not to mention a little too old as well ;).
- dmourati 8y agoAt least you are honest about your ageism.
- scarface74 8y agoHow is that ageism? There is nothing stopping a person in their mid 40s from keeping thier eye on the ball, not getting complacent at a company, and not letting thier skills atrophy. I’m in my mid 40s and also know plenty of other developers my age who decided not to go anywhere near management so we had to keep our skills up to date. There is nothing sadder than seeing an older developer who stayed at a company for 15 years, get laid off and spend months not being able to get a job. On the other hand, I have a former manager who retired from the military in the late 80s, started out as a Powerbuilder Developer and became a manager at a .Net shop. When his kids moved out, he “self demoted”, got a job as a developer and is now doing .Net and React. He is at least in his mid 50s.
- dunpeal 8y agoThis isn't my ageism, I personally love experienced candidates. This is a reflection of the reality in SV, where most employers aren't like my current one.
- scarface74 8y agoFirst of all, what "cost of switching jobs" are you referring to? Besides the job hopping stench, there are other very real cost to switching jobs. - there are three “levers of power” in an organization - relationship, expert, and role power - in that order. At any job, especially at more senior levels, having the right relationships with the right people means for example the difference between you having to send a ticket to the IT department and wait in a queue to get those test VMs you need and being able to send a Skype message, they set everything up for you right then and then you submit the ticket as a formality. - Expert power and developing a reputation for knowing what you’re doing is the difference, especially at a small company, is the difference between having to wait for the “IT department” (or the outsourced Amazon support people that your company hired), and being given admin access to AWS so you can do it yourself. When people know you and trust you, you can punch above your role. You lose that at a new company and it is a “cost” to get it back. Also, at a company I’ve been at for three years, I know where most of the bodies are buried. I’ve buried a lot of them or know who did and I know the technology. I spend the first 6-12 months at a new company working extra just to ramp up on the company’s code base, and because I’m usually changing jobs to learn new-to-me technology, learning the tech stack. That extra time has a “cost” on my personal and family life.