People change jobs at the drop of a hat for reasons that are well within companies' control. It's not like people want all that stress and hassle, they do it because they're incentivized to do so.

Sometimes. Other times they are just sampling what's out there to see what suits them the best, or playing the comp boost game until even their best face forward isn't able to garner a higher offer.

The former happens, I’m not convinced that it’s a common occurrence. Changing jobs is genuinely stressful, I don’t think people do it lightly. The latter is usually something the company could fix, but won’t.

Even still, it’s obvious that comp alone isn’t enough to retain employees. Even FAANG companies, which pay extremely well, have pretty low retention numbers. Facebook does best here, at an average job length of only 2.02 years. If comp was enough, people would stay there longer. This implies that people are changing jobs for reasons aside from “this other company will pay more”.

It’s not just about absolute comp. If google will pay you more, then maybe you leave facebook. That’s not because google pays more than facebook, just that it’s easier to get a “promotion” by taking a hire role elsewhere than it is to get an actual promotion. It doesn’t mean you don’t pay market wages, but it does mean you don’t pay that person their market wage.

That’s exactly something companies have under their control. If people are leaving because it is seen as the sure route to a promotion, then perhaps providing clear advancement opportunities internal would reduce this phenomenon and help keep your high performing talent.

The present employer and prospective employer have a different perspective on the individual. It's entirely possible, and indeed somewhat common, that individuals are hired to levels to new employers beyond what they would be able to justify promotion at their existing employer. The only way to eliminate this is prolonged and comprehensive interview process, which is what TFA is railing against.

I disagree that this is the only way to eliminate this problem. Companies could loosen promotion criteria to be more in line with what external candidates bring. Ultimately the cost of lost knowledge and backfilling is quite high, and could easily justify faster promotion cycles on a monetary basis alone.

Even if some of them genuinely get promoted before they’re ready and wash out, you’re not really that much worse off than if you’d lost them before. Besides, there’s always the risk that your new hire is unprepared too, which is a much harder thing to quantify.

I have no idea why people think this. I've trained all sorts of people on all sorts of new things and it was always worth it.

Sometimes they leave. By that time they've used what they've learned and passed some of it on to the next person.

Incredible that you're the first person I've seen mention a second-order effect in this conversation.

I don't know what it is but I feel like people have forgotten just like basic truths about how humans work. Maybe it's because managerialism has infected everything.

Then give them a reason to stay (note: it's also not all about money)

Whatever you offer, including non-monetary, someone else can offer and then also spend their training budget on higher comp.

That works if this is a "one time game" (game theory) as opposed to a repeated game.

If an employer does do training, it means they'll probably continue to do more training over time, which helps the employee become more valuable.

If the employer doesn't do training, yes they may be able to allocate the training budget to salary, but they are not going to spend anything training you or letting you work on projects to increase your skills while you're there, unless they absolutely must.

I think people also have some human perception of how they're being treated, and prefer to work for people that invest in them.

Not necessarily. It's not easy to find a boss who you genuinely trust to consider your best interests, for example.

Out of curiosity, what sort of "non-monetary" benefits were you thinking about? There's usually not a reliable way to turn (small amounts of) money into the sorts of things that really build loyalty.

Is this supposed to be a rebuttal? I don't see a problem here.

Yes. Spending money and then not recouping is a losing strategy.

So the answer is to not spend the money at all? How much are you costing your company by putting candidates through 8 hours of interviews only to reject them. Rinse. Repeat.

All the while, productivity suffers as the remaining team falls further behind due to short-staffing and being pulled away from their real jobs to interview.

Professions mandate training minimums per year in order to maintain credentialled status. They're low, sure, but they at least create a need for ongoing professional education.

Working at a company that doesnt pay for my training while my skills fall behond is a loosing strategy.

I literally said it isn't all about money. Most people leave because managers[0]

> In general, people leave their jobs because they don’t like their boss, don’t see opportunities for promotion or growth, or are offered a better gig (and often higher pay); these reasons have held steady for years.

And if it is about money, then this is called paying competitively.

But lastly, recognize that if everyone is training employees you're still not really losing out unless you're only hiring entry level employees. Sure, you might be training someone that leaves, but so does your competitor. But if you're only hiring junior engineers then you're probably doing something wrong that's much bigger.

[0] https://hbr.org/2016/09/why-people-quit-their-jobs

Every process has pros and cons, you have to weigh the net benefits.

That works when you're employing a bunch of Wordpress monkeys who do nothing all day but mess with CSS and install plugins. Not so much when you've got a mature SAAS product, parts of the system are tricky to work with, and stakeholders are breathing down your neck to implement new features so you don't have time to cross-train your teams.

Losing people who are experts within the domain of the software they're maintaining because you refuse to invest in them is going to cost you thousands of dollars... the only question is whether that's tens or hundreds times that amount... and in some cases it can cause you to lose your entire business.

I’m friends with a few “Wordpress monkeys,” as well as some people who “do nothing but mess with css all day.” That was extremely condescending and dismissive

Yep. Apprenticeships solved this problem in the past (and of course created many others). Actually it’s almost a fun little exercise in economics.

Basically there’s two types of efficiency, investment efficiency and allocative efficiency. (There may also be other types I don’t know about.)

Investment efficiency means people are incentivized to make positive-expected-value investments. Think about how people are incentivized to invest in their house, e.g. preventative maintenance, because if the expected value is positive then they will recoup that value when they sell the house. If you’re renting you don’t have this with respect to where you live - water damage or no, not really the renter’s problem. Investment efficiency is maximized by private property, where you know that no one will take your property without your consent.

Allocative efficiency means things go to whoever is willing/able to pay the most for them. Renting does have this property - if both of us want to rent a house, and I’m willing to pay more, in most cases I’ll end up getting the house. This is why gentrification can cause displacement - when wealthier people come into a city and are able to outbid the current renters, they win and the current renters lose. Allocative efficiency is maximized by auctions and things like them, where the good goes to whoever is willing to pay the most.

Bringing it back to your comment, job training isn’t worth it because our careers as programmers are dominated by allocative efficiency, not investment efficiency. If you can train a programmer create $50,000/year more value in general (i.e. it’s not training that would only be useful to your company), they can now get paid about that much more from any of your competitors, and you will have to pay them about that much more to stop them from leaving. So you gain nothing from giving them general-skills training.

Another way of solving this problem is with sectoral bargaining. If you have a sector-wide union, they can make all companies start training simultaneously, or assume some of the costs themselves. It’s a win-win for the industry and for the programmers, but it doesn’t happen nearly as much as it could because of that coordination problem.

>Yep. Apprenticeships solved this problem in the past (and of course created many others). Actually it’s almost a fun little exercise in economics.

But it makes Reginald the investor angry that his ROI isn't exactly 20% each quarter, so they jettison apprenticeships and start cooking the books to make that possible.

So, in this hypothetical, the sector-wide union is preventing individuals who learn to create an additional $50K/yr in value from realizing the increase in pay which would otherwise accrue to them?

In return for wasting training on employees that will leave for other companies, the company is getting its employees trained for free by other companies in the same way. In aggregate everyone wins because employees now get training.

The union is ensuring that no company can ruin it for everyone.

Yeah, or at least they aren’t able to capture the entire $50k/year in value. It kind of sounds bad but it’s a trade and there has to be something in it for both sides for it to happen.

You say that and I’ve seen several companies in practice echo what you’re saying. However, I fail to understand why they don’t simply make better use of contracts and probationary periods to solve that specific problem.

The problem of a probationary period is that it pushes all the risk to the employee.

While I agree interviewing has gotten ridiculous with all the leetcoding and ten rounds of interviews and FAANG cargo-culting and whatnot, one small advantage - assuming I'm not desperate for a paycheck - is that it gives me, as a prospective employee, time to consider and withdraw my application if I see too many red flags or I just prefer the devil I know.

A short interview process with a probation period on the other hand is a big roll of the dice. Maybe I'm not able to ramp up on time, or make a silly mistake due to unfamiliarity with the codebase or underlying business logic. Maybe I don't get on with the team or manager. Maybe I'm going to be dumped into a doomed death march project on day 1. I could find myself unemployed a month later with an embarrassing gap in the resume. Perhaps on the other hand a better interview process (not longer, just have properly trained people and constantly improve the process with feedback) would save us all that pain.

In a world where short, high-risk interviews dominated, you could just go roll the dice again. It would be a negative signal (why is @foo interviewing after only 60 days?), but nowhere near as bad as “why is @foo still interviewing after 6 months in this job market?!”.

Contacts in what sense?

Probationary periods could work but it's a coordination problem. Such periods are the norm in Europe (coz it's very hard to fire someone) but for an at-will place like the US, given that the industry doesn't really do probationary periods in general, any employer who starts doing it would be at a disadvantage.

*contracts

In the sense of offering signing bonuses for term lengths that get repo’d if the contract length is broken.

I think GP meant "reference check"

I meant “contract” but you’ve brought up another good idea.