Explaining tech's notion of talent scarcity | Hacker News Reader

Explaining tech's notion of talent scarcity

nadia.xyz

74 pointsdcreager38 comments

This doesn't say much about "talent scarcity", it's more along the lines of how different managerial styles map to different talent "markets".

In reality, there is no talent scarcity, there's just poor management that has optimized for executive salaries and middle-managerial busy work. Paying your managers/executives more than your engineers, forcing wfo, open office plans, agile... all chase away top talent of which there is plenty. The shortage is actually of healthy, creative, non-toxic, fairly compensated work places. Deliver that and the talent is easy to hire and retain.

I think a lot of the friction is upper-middle or middle-but-anxiously-trying-to-claim-upper-middle (in a class sense—income's related to, but not identical to, class, in this analysis, following Fussell) management not wanting their reports to have equal or greater standing, in social status terms.

Probably this treating programmers as a status-threat dates to the merging of high-status analyst and low-status programmer into a single role, way back in the 80s—there was little threat with a few upper-middle or near-upper-middle analysts running around, but increase the total number of people in programming and merge "analyst" and "programmer" and now there are too many, and no-one to be "under" them in status, and what's the point of status if you don't have inferior people to lord it over? If there's nobody you can credibly boss around or tell to take care of some bullshit for you, and expect them to just say "yes" and do it without pushing back? Nobody to do the actual work while you're "working" at the golf course or goofing off in your private office! Clearly, that would be a threat—why, they might all get dragged down to bottom-tier status, if there's no one under them!

One bit of corpo-speak that I have seen cropping up to address this is relabeling "managed by" as "supported by", as in, "team X will be supported by person Y." It may not work overnight, but I would like to see a shift in the culture of a manager "managing" like a king manages his kingdom and the peasants within, to "supporting", like a coordinator. But I fear that we will fall back to the ape mentality of hierarchies and end up twisting the words to fit our predetermined outcomes.

I’m a program manager my role is very close to this. People are confused when I give them my title and explain that I don’t have direct reports, but I “manage” our team about as much our director does.

That means handling administrative tasks, organizing work, stakeholder management, or any of the other million tasks involved in running a successful team that developers shouldn’t need to worry about. It’s very easy to conflate prioritizing and planning with telling people what to do, and sadly in many orgs that is the case. However, those tasks and everything else falling under the “management” umbrella should not imply a hierarchy.

It’s just a different type of work to be done by people with different skill sets. I don’t know if that reality will ever be embraced by my fellow business bros, but that simple mindset has made me a better manager than any class, book, or course I’ve experienced.

I read the concept recently of a people manager vs. a process manager. A program manager from my experience is a bit of both but mostly process.

If you don't mind being part of the "support staff" you can have the job!

You can't change human nature. Hierarchy is inevitable. All you'll do is replace explicit hierarchy for a covert implicit hierarchy, because most people are natural minions who will look for someone to take charge.

I note that there two pervading ethea on HN, both exemplified in the parent comment:

1) Poor management (broadly defined) creates misaligned incentives, busy work, etc. that "gets in the way" of good talent executing well.

2) Management, who "doesn't really do anything," is overpaid relative to the IC's who "Do the work."

These two ideas, though, are in contradiction. Management deserves higher pay exactly because poor management can destroy value so quickly and easily. The problem is identifying, within the "company system," good management from bad management.

There are a lot of other factors at play:

A lot of less experienced employees don’t understand what managers do. There is also a certain personality I’d call “the geeky engineer” who thinks the same way.

Some things that are actually signs of an effective manager are hated by employees. Some individual employees for example need micromanagement or frequent manager involvement to make sure they don’t go off track, get stuck without telling anybody, or blow a time budget on something unimportant (eg a task is scoped at 2 weeks and they spend the entire first week playing with something speculative that may help).

A lot of managers are both highly paid and truly bad. I think a lot of businesses don’t do as good of a job vetting EM hires as they do SWEs the EM will manage. And I don’t mean that the EM needs to be Staff-level necessarily.

Even a good manager is incentivized to do things that are globally suboptimal (empire building, overextending) for career growth. This is also true of ICs though.

Personally, I think the biggest problem is the EM hiring bar though. Where I work has high standards and is very selective when recruiting ICs, but it seems like all that goes out the window for external EMs. I do think EMs deserve high pay and to be incentivized to do good work, but that requires a hiring process that better distinguishes between good and bad EMs.

If your employee is blowing their time budget they are most likely either:

a) a bad employee who you should not try to micro manage to make effective but should just replace with a better employee

b) unaware of the context around the task and does not have aligned motivations with others to get this done quickly

If b - then a good manager should do a better job of figuring out how to develop the intrinsic motivation of the employee and find ways to align the incentives and motivations of management and employees better.

Trying to micro-manage to avoid employees doing things you don't want is putting a band-aid on the problem. In fact it's even worse because you're creating a bad work environment that will attract more sub-par employees who accept micro-management and will scare away the high-performing employees you so desperately desire. Eventually you have a bunch of micro-managing managers who themselves are being micro-managed and your company bloats to several multiples in size of what is actually required and efficiency is spiraling out of control.

There is not necessarily a dichotomy between good and bad employees, and I’m certainly not saying every engineer should be micromanaged, but IME certain engineers require it to be able to perform well.

I think it should be applied on a case by case basis when necessary, and I’ve seen effective managers do this in the past. Anyway, the type of employee who requires this treatment often is precisely the one that least enjoys it, which is my original point.

How about C - someone who sees a lot more at hand than a specific problem being worked and goes on tangents fixing issues management isn't aware of or doesn't understand. It's career limiting to operate this way, but is long-term better for the company.

These people deserve a better company, and should seek one. Or start their own.

> A lot of managers are both highly paid and truly bad. I think a lot of businesses don’t do as good of a job vetting EM hires as they do SWEs the EM will manage.

From my vantage point as an engineer it also seems like what qualifies as "good management" is hard to pin down because not only is it difficult to quantify in the first place, but it also varies between teams and organizations… a management style that worked in one team and company might fly like a rock elsewhere. If that's true, it'd make vetting managers challenging even if the business in question is trying to vet them as well as they do engineers.

The problem, IMO, is the muddying of the managerial role, such that they now do 2 separate roles, one of which is overpaid, the other not. The computer all but killed the secretary, and the remaining (computer-assisted) secretarial duties got split between the ICs and management. Because you can only dump so much secretarial work onto your ICs before they stop being able to produce anything, the managerial level got a bulk of the secretarial work, which in turn prompted more hiring at the managerial level. On top of that, it represented a flip-flop in power dichotomy, because the managers still have their managerial duties/powers: the people doing a lot of the secretarial work now have the power to fire the people doing the product work. Combined with people's self-interest in protecting and building their own positions, this leads to secretarial work bloat, and thus even more managerial bloat, and an increase in secretarial workload on the ICs.

My pitch for a the ideal corporate structure: bring back the secretaries. Re-introduce the division between your secretarial work and your managerial work, and shrink the number of managers by a significant fraction. Ideally, the result is A) fewer managers means more budget for hiring better mangers B) less secretarial work among your ICs and managers boosting both productivity and job satisfaction C) flatter management hierarchy means less information loss between ICs and the top decision makers and D) the people doing secretarial work no longer have the power to shunt that work off onto the ICs.

I'm curious about this. What kind of secretarial work do you have in mind, for a typical tech company? What fraction of time do you estimate it takes? Meeting scheduling is automated today. What else?

Document management and user access control are two I can see.

> 1) Poor management (broadly defined) creates misaligned incentives, busy work, etc. that "gets in the way" of good talent executing well.

> 2) Management, who "doesn't really do anything," is overpaid relative to the IC's who "Do the work."

> These two ideas, though, are in contradiction

I don't see them in contradiction. They're also neither mutually inclusive nor mutually exclusive. They are separate observations and both may be valid or invalid for a given set of management.

I happen to think that there's a lot of good "management" but not a lot of good executive teams. Sure there's definitely a lot of bad management too, but IMO there's a far bigger pie taken from poor (or outright employee-hostile) C- or S-level executives.

I said that there is no "talent scarcity" and anyone claiming that there is can count themselves among the rank of bad management. If management can't attract and retain talent, it's by definition bad management.

no, you are missing the usary aspect of market forces, where getting more talent giving fewer privileges for $PAY is the stated objective. People who practice skilled work are blind to business motives to abuse skilled work specifically, because "why would you do that." The commoditization of replaceable "skilled" work roles, with increasing oversight and metrics, is absolutely a factor in high tech.

I think the only solution is more managers to manage the existing managers

I wholeheartedly concur. Thanks for putting it into writing.

A similar debate exists within the professional military community about the appropriate place to give special operations forces within the broader force. There is a concern, especially in the United States, where special forces have accrued a growing share of resources and personnel since the beginning of the global war on terror, that such growth is draining regular forces of all their top personnel and thus hollowing the regular units they rely on.

https://cco.ndu.edu/News/Article/1020184/the-limits-of-speci...

https://ndupress.ndu.edu/JFQ/Joint-Force-Quarterly-108/Artic...

https://www.reddit.com/r/WarCollege/comments/gahsnv/does_hav...

In careers, positively-contributing people are motivated by three things: money, team, and perceived importance of work.

What about 0 and negatively contributing?

One of the interesting things in this article is how OKRs are great for a "Pareto-distribution" culture, where performance can be measured across employees, but not a "linchpin" culture, where employees are exceptionally good at a small handful of things and not necessarily great at doing whatever the company needs. I've frequently seen those two types of cultures clash at the seams between technology and business.

I don't get it. How are the distribution graphs connected to the text? What is distributed over? I guess the people of the company? What is the y-axis? The amount of talent in a person? The amount of special talent in a person? I hate pseudo-scientific explanations like that.

It's not you. The article relies on a bullshit method of making its simple point, deliberately overcomplicating it for the purpose of signalling.

Tech's notion of talent scarcity:

1. "We can't find enough cheap labor to work at our awful jobs."

2. "We're doing everything we can possibly do attract talent (except for raising wages, which is of course impossible.)"

3. "Never mind, we don't need human workers anyway."

A competent overview of the relative ‘bins’ of labor used to power large firms around the world. The use of distributions really tickles my brain and helps understand not only where I work but also why some people love the places the work (because it fits into their style and desired work environment too).

I see companies ignoring or actively scaring away talent left and right.

This has been sitting with me for a couple hours now. The impact was semi subtle at first but has really grown as I keep reflecting how much better it fits my personal view.

The Bimodal model really reflects a lived truth, that explains things differently than Normal or Pareto curves. Some people aren't just higher output but are fulcrums of the business, are core drivers.

Our malicious website filters prevent access to this domain - is it legit ?

Your filters might have rules about .xyz. Otherwise it's legit, the only connections it makes are to google tag manager and google fonts, loads ~220kb, and works without javascript enabled. Hard to be malicious without javascript!

But for any domain you're unsure about, you should likely try a sandbox to get a definitive answer for yourself.

More likely: your incompetent website filters prevent access to this domain.

Probably just because of the .xyz domain? But yeah, per the other comment, it seems fine.

It looks fine to me.

Yes.

Reply on news.ycombinator.com