Sure, it sounds that way, if we assume that negotiating skill and job performance are tightly correlated. But why should we assume that?

Above a certain threshold of talent, there is no need to negotiate at all.

If the two traits are just randomly assorted, then sure, they'll rarely overlap at what is a statistically random rate.

But, if we can assume that the "rare talent" is anything other than straight out of college, they will have discovered that they are a rare talent, be aware of it themselves, and barring some sort of crippling neurosis, want to cash in on it. They might not even be skilled at negotiating... they'll just be aware that they have leverage.

We might even phrase the problem that way... that it's not so important that they have "negotiating skill", as it is that they have "leverage awareness".

If it were that simple why would anyone need tips for salary negotiation? Even in pro sports, where the skills are extremely elite, all the performance stats and salaries are public, and the athletes have agents in their corner, the correlation between pay and performance is far from perfect, and negotiation mistakes can lead to players leaving huge sums on the table.

If you literally wrote the book on the subject, or something similar, they'll probably just create a job role for you.

But that said, not every company is trying to hire the best people. As you note, they're expensive.

You'd think that, and it's probably even the smart strategy for the employer... But I remember what the guy who wrote homebrew said about his Google interview process.

Once an HR culture develops, anything sane or intelligent probably goes out the window.

Writing a popular app is hardly the same as being the top authority in a field.

I guarantee when Google hires someone like Ken Thompson, they do not get the normal Google hiring experience.

A popular app? I hardly think homebrew is on the same level as candy crush.

Yeah, Candy Crush is harder to reproduce.

I also doubt that Kenny was having to do daily scrums and explain why the web ad wootigulizer algorithm hadn't had any progress since Monday. Long since in executive management, is my guess.

Let me just preface that inverting a binary tree.. Maybe they wanted a particular answer and it was an off day.

That said, the author of Homebrew would be the top authority in the field of OSX package management yeah? I mean, Apple hasn't even bothered to create a competitor and yet every tech bro at Google using a MBP is going to need to install and use it..

IIRC the homebrew guy ended up going to Apple and worked on... Swift Package Manager. Or maybe I'm misremembering things

Maybe the homebrew guy isn't a good hire?

There's an old Dilbert comic that said something like "We want to hire engineers who are really good at what they do, but cannot compare two salary numbers and identify which one is larger"

[delayed]

The problem isn't that; it's all the information you don't have such as "how much are they theoretically willing to pay me," "who else would hire me," "how hard can I push before they just pull the offer," "do they have other candidates they liked," and so on. On top of that, lots of people in this career are not smooth talkers and they're risk-averse.

You get what you pay for.

If there is a one off person who can command a higher salary, its unlikely they alone will make a huge difference to the company anyway.

If there are a lot of people who could command a higher salary in that role, the pay is too low.

Overall it seems like a pretty good system. You could argue that the current system favors people who are good at negotiating, and not necessarily more skilled workers.

There are lots of people who aren't choosing jobs based solely on which one pays the most. I have absolutely taken positions that paid much less than I can command elsewhere because the companies in question offered me nonmonetary benefits that were worth more to me than money.

Whenever an employer talks about industry median pay, I ask if they want industry median work, or industry median talent and remind them that half of all will perform better.

Not if everybody in the entire industry is paid the same. :)

Not having a band is surprising at Fortune 100. I'm guessing the talent retention is still done through bonus and LTI.

Oh, there are band levels (at least, assuming we used band levels in the same way you mean), but everyone in that band level must make the same amount of money. And the band levels are fixed. It's not like talent acquisition can magically make up a new band level to give someone more money.

If everyone is making the same amount, then there is no band. Band = Range with midpoint. For example, Sr. Software engineer's pay band could be $120,000 - $180,000 with $150,000 roughly being the midpoint, with most people making around the midpoint mark. Now I'm curious what company it is :-)

On the other hand, they probably spend significantly less time bickering about their respective salaries.

IME, talented people can spend an absurd proportion of their time bickering about how ${other_talented_person} makes ${trivial_number}/year more than they do, all because they're ${some_trait}.

Overall, I could imagine this filter being both good and bad. I'm not confident whether it nets out good or bad, but you can't argue it's not a dramatically simpler system, and simple systems are often surprisingly better better. :shrug: