I think the math is more complicated than this makes it seem.

The missing bits are that many employees will stay even though they are underpaid and many people will still leave even if they are given retention raises.

If an employer gives a raise, that is a reoccurring expense they might not have had to pay; it might decrease the chance an employee leave from 25% to 10%, which means it is an expense they didn't have to pay 85% of the time (75% they would have stayed anyway and 10% chance they will leave anyway). 100% of the time the company will have to pay the expense, though.

The few people you do retain from the raises are likely going to be people who will leave for a better job at some point anyway, even with the raise.

After a number of years of giving retention raises, you will likely lose most of the people you retained for a short time with the raise, while you continue to pay the increased salary to people who would have stayed anyway. Over a longer period of time, it becomes almost pure cost without a ton of benefit.

There is obviously a balance here. You want to give the right raises of the right amount to retain your best talent without spending more money than you have to. If the pay is bad enough, you will lose all your best talent.

A good HR department would know these things. They will know industry standard retention rates and know if the company is above or below the average. They will talk with managers to assess the quality of the people leaving and why, and adjust raises and bonuses as needed.

While this sucks as employees, it isn't irrational.

Treat average employee tenure at the company as a KPI.

I've worked at companies that without a question raise all employees to market like clockwork.

I've also worked at companies that skimp on raises.

The automatic raise company had tenures in the range of ~10 years, which meant incredible institutional knowledge and contact networks. Consequently, things that would be impossible at other companies were trivial there.

Long tenures can cause their own issues (specifically: not firing people who need to be fired), but are a huge tailwind to productivity, especially at the architect level.

And side observation. If your company is only profitable because you're underpaying your employees, build a different company. If your company is more profitable because you're underpaying employees... you're effectively giving the missing employee salary to shareholders. Gross. Employees are the ones doing the work.

Perfectly fine to make whatever profit the business/market supports.

But if a company is making profits while underpaying their employees relative to market rates, yes, that's backwards.

It's almost like companies shouldn't try to win a war on two fronts.

Instead of having shareholders and customers, just have customers. Then, yes, all the profit would go to the workers/owners.

> They’re addicted to the next shiny thing—the new hire who “must be better” because they’re fresh talent.

Here's another take: Job searching is painful for employees. Employees must be offered an incentive to go through it (more money for new hires) and are willing to accept some loss to not go through it (being paid under market value for existing employees).

That's why i don't. Especially considering unknowns. If i'm not actively wanting to leave my job i probably generally like it, and who knows what the next job could be like.

Leaving my current job might mean just one new job or a half dozen until i find a good place again.

Me too. Current job has a lot of positives but of course some negatives. New job could be better, but probably a wash overall.

The pain of job searching varies. The more competent you are, the more opportunities fall out of sky (recruiters and personal network). The more underpaid you are, the better those opportunities look.

If it's hard to find a job, then, well, yeah, you should expect a mediocre and stagnant salary.

The differential is to avoid hostaging.

Recruiting is n=1: overspend b/c the need is immediate and well-defined.

Retention is n=M: underspend b/c the risk of everyone threatening to leave once someone successfully threatens to leave and gets a pay boost is incalculable.

My guess is that companies do spend extra on retention when confronted with the thread of someone leaving, so long as that threat is quiet and implicit.

Likewise, incoming salaries might be suppressed to the extent that overpaying if published could stir resentment.

Companies spend serious commissions to salespeople who cause customer churn from overselling, while underspending on the quality and security that would keep the customers in place.

End of the day, hunting is more attractive than tending.

I think it's stock market mindset. Investors love "growth", so "we had #### new subscribers this month" is sexier than "we retained 300 million subscribers", even if "new subscribers" was "people who haven't used us for a year" in fine print.

They might "overspend" compared to retention, but I don't buy that recruiting is actually valued and over-budgeted by most companies. It's a cost center and there are very few direct incentives & feedback mechanisms for having a good process. Do people actually see managers getting fired for failing at recruiting?

Companies do not care. They treat employees like tissue paper. Sneeze, throw out, pull out another one.

It’s similar to rental apartments. They think the challenge of changing jobs/apartment is bigger than the improvement of salary/rent. They’ll difference is the smaller salary increase or the rent increase is maximized to help their finances not yours.