On the Law of Diminishing Specialization
calnewport.com
calnewport.com
The trend in the new economy is for a large percentage of people to be discarded like rubbish, while increasingly insane workloads are placed on the remainder. Not only is this bad for all the individuals involved and for the health of society, but OP makes the point that it's actually bad for the bottom line too. Executives are rewarded for doing things this way because it superficially looks like they are saving money, but the resulting productivity cut ends up reducing profitability. A more sane and equitable distribution of workload, would be better by every metric including profit.
This trend started over a century ago. As mass production took over, skilled labor (artisans and tradesmen) were replaced with people who may as well have been machines (for all their employers cared). Scientific management led to a hyper-micromanaged workforce (at least as a goal) where each worker was instructed in the most efficient way to do their job and weren't expected to deviate.
Interestingly, Taylor saw some of the same things Deming did but came to very different conclusions about how to resolve the issues. Taylor effectively worked to disempower workers (failing to recognize that this was also what led to their demotivation and poor performance) while Deming worked to empower workers (recognizing that this led to pride in their work and better performance).
The problem has been that, while Taylorism has been discredited for nearly a century, his influence remains. It moved from the factory floor into the office.
"If only you devs would do everything right, we could always ship on time with zero defects." -- An almost verbatim quote from a former supervisor, not ironically
Many people in management roles are woefully underprepared for it. They operate under the same flawed belief as Taylor: If only the stupid employees would do precisely what the smart people wrote down in the work instructions (NB: often not actually written down, or not matching reality if written down) then this company would be awesome. Why is it so hard to do a travel voucher (ignoring that the employee travels once every 2-3 years and the voucher forms have changed since the last time)?
> Executives are rewarded for doing things this way because it superficially looks like they are saving money, but the resulting productivity cut ends up reducing profitability.
From understanding Lean, there are costs you can cut (things which aren't value added), and costs you oughtn't cut (or shouldn't cut much). But it's easy to recognize when someone is doing work which directly adds value (writing a line of code, properly aligning the mold and the sheet metal). It's hard to recognize indirect value and costs. I had (but removed, it got long) an analysis of how much we lose in my office to time spent managing travel each year. The swing would be from losing over $100k/year to making over $100k just by hiring one full-time travel specialist (we don't charge the customers for the time spent dealing with travel agents so the engineers' time is lost revenue, a single in-house travel specialist could handle the local office with ease and free up those engineer-hours to be charged to the customer for actual work).
The Dev/Ops interface is one with very little information hiding, little responsibility segregation, very little outcome independence, and nearly no knowledge differentiation.
The researcher/admin interface separates things with complete opposite qualities on each of those dimensions.
Both situations have a jack-of-all-trades problem lurking behind them.
I worked for a small company (5 or so employees). We had an administrator who was a former VP of a very large bank. I don't know how much we paid her, but she was worth more than any other employee, in my experience. She made that place run.
I suspect that one very good administrative worker is worth their weight in gold, and has the ability to make 10x-ers out of otherwise normal employees. I'd suggest paying for really good admins, rather than viewing them as low-skilled, low-wage commodity workers.
It may well be statistically bad for the bottom line, but it's much easier for the day-to-day. Planning things is hard, decomposing work is harder still, and making sure that the things you build are architected and organized in such a way that they can be planned and decomposed is the hardest yet; it's the same reason parallel programming is hard, I think (this, granted, all from a software perspective).
So even if yes, it may be biting people in the long term, the short-termism is that it's way easier to get things done because you (as the exec) just ask one person to do the entire project, and they maintain all that complexity in their head, never needing to translate it out into shallower, noisier channels, and the 'difficulty overhead' of sound planning, architecture, management, and communication is avoided.
The "I'll just do it myself" culture is really hard to escape from, too, because I think we feel more productive in it, even if we aren't (compared to an organization that effectively staffs work out, and has the staff to accommodate). Intuitively, you're doing all this work and your fingers are constantly flying across the keyboard. You've got so many commits in today. You feel a lot more powerful than you do writing up issues and stories, and asking for statuses. And at the same time, it's a lot easier than writing up issues and stories (ones that will be genuinely useful as planning and work items, at any rate). It's part of why hero-efforts remain pervasive, I think. And it's a kind of sunk-cost fallacy at the end of the day; you're so deep in it already that you may as well just put this one last little bit of effort in and finish it yourself, rather than waste all the time documenting and decomposing and planning so that you can better facilitate parallel, distributed work.
Paying somone minimum wage to just keep that from happening across say 15 conference rooms is a likely a net win even if they did nothing else. The issue is this is an invisible cost where cutting that person looks like real savings.
We have this idea that everyone needs to be actively working all day when there is such a huge discrepancy in pay scales that it’s not that important.
People want to justify their purpose in an org, so they will create approval or audit processes where they aren't needed.
With more people, it takes more effort for knowledge to spread, so you either spend that effort (which reduces efficiency), or you have people doing their jobs without information that could make them more efficient (which reduces efficiency).
I've heard arguments that adding administrative staff is worse than adding other sorts of staff, but that is not as immediately obvious to me.
5 people can produce 10 widgets in a year.
10 people can product 15 widgets in a year.
Are the 10 people less efficient than the 5?
If you need 15 widgets, the team of 5 has 0% efficiency.
A Honda Fit is clearly a more fuel efficient car than an F1 racer. This remains true whether or not the problem you are solving is "win an F1 race," which would necessitate using the latter.
If I hire 15 people, I can produce 20 widgets
If I hire 15 people + 5 lower-skill workers, I can produce 30 widgets.
The problem we face is visibility. The question becomes how is it that low-skilled workers can create more efficiency, they themselves don't do anything. The reality is that high-skilled workers are now free to get creative vs having to do the annoying tasks the low skilled workers just took off their shoulders. So _indirectly_ the low-skilled workers produced 10 widgets, but really, everyone was just working at a good production level.
People being less busy lets them be creative.
If you just add more administrative stuff together with other people and specialists that already have n*(n-1) links between them, this will be a problem due to obvious reasons.
However, if you actually spend time to work out a process and organization that will not require additional each-to-each communication and make most of the communication remain in close groups, you may avert this. You can call this a human refactoring, if you will - very similar idea to separate functionality into small, encapsulated classes with smalles possible public interface instead of one huge g-d class.
I dont think this is an exception to my rule; you can reduce the overhead significantly with good process, but any process that does not perfectly predict who needs to know what information will incur some per-person overhead as a team grows.
Perhaps you are arguing that this overhead can be reduced to the point of being close enough to zero to not matter? If so, that would be very impressive, and conflicts with my experience.
Small ships can't handle big cargo.
And if you're that administrator, you probably don't need to justify your purpose, you should probably ask for a raise.
I'm not sure where else economies of scale are supposed to come from.
Why don't more executives hire really good programmers and analysts to be their assistants? If a tech CEO is earning $5m a year who gives a damn if their assistant is earning $500k? Right, like why not just pay for the boost in automation and understanding of abstract concepts to maximize the communication-time to useful-actions ratio between assistant and CEO?
You might be thinking of startup and small tech company CEOs that really understand the nuts and bolts of a software product, but they're not making $5M/year.
They do. Usually they have a title like "Chief of Staff" or "Speechwriter" or "Associate." My friend occupied just such a role after working in investment banking and private equity. He was paid in the low-mid six figures and ended up doing tons of research, analysis and writing for the CEO of one of the most successful private equity firms. The position was stepping stone to Harvard Business School and nobody stayed in the role for much more than a year or two, so they are constantly rotating in fresh, hungry and smart people.
I like the idea of adding automation to the mix, but realistically, if there is a task the CEO wants automated, he/she can tap any engineering manager and pull their team onto that project. The value add would be automation of tasks for the assistant rather than automation that benefits the CEO directly.
In a startup it may be faster to do everything yourself. Once you operate at scale, every hour of low value activity sucks away an hour of high value activity. I don’t respect folks who are incompetent without their admins, but I also disrespect managers too stuck in the weeds to lead.
Yet this is exactly what the article is proposing.
If you don't do, say, Powerpoint, with enough frequency, you are effectively incompetent at it. Insourcing this to an admin practically guarantees you will be incompetent at it.
Steve Jobs had Apple create iPreach (Keynote) because Powerpoint wasted so much of his time ...
The article is talking about tasks that actually are necessary, because they are still being done. But instead of being done well, by a person specialized in that task, they're being done less well by people who are specialized in entirely different tasks, taking away the time they can spend on the tasks that they themselves are specialized in.
> In other words, the author claims that when we speak of “bullshit jobs,” we’re really just talking about the postindustrial equivalent of factory-line workers, those with the unenviable fate of having to carry out the repetitive, mind-numbingly boring but still very necessary tasks required to manage increasingly complicated processes of production. As robots replace the factory workers, these are increasingly the only jobs left. (This position is sometimes combined with a rather condescending argument about self-importance: if so many people feel their jobs are useless, it’s really because today’s educated workforce is full of philosophy or Renaissance literature majors who believe they are cut out for better things. They consider being a mere cog in administrative machinery beneath their dignity.)
> Final Working Definition: a bullshit job is a form of paid employment that is so completely pointless, unnecessary, or pernicious that even the employee cannot justify its existence even though, as part of the conditions of employment, the employee feels obliged to pretend that this is not the case.
[1]: https://www.economist.com/free-exchange/2013/08/21/on-bullsh...
Moreover, I'd guess the overhead of communication probably means it is more efficient for a marketing director to make his own presentation and charts. After all, he is the only one that knows what he wants to present, and what data should be in the charts. If I had to communicate that, I'd probably make a crude powerpoint and say "Like that!".
The computer systems caused secretaries to get fired while managers get to do secretarial tasks of feeding computer systems with data... There are probably much better uses of actual managers.
The thing is, managers employ person to person skills and unstructured data handling. Forcing the manager to structure their output so that the computer / bureaucracy can handle it is a waste of time. If you want this kind of accountability, hire an accountant or secretary.
Welcome lean processes and reporting based on word and thin paper trails only for very important or regulatory matters.
Or change the computer system to accept unstructured input.
If something is offensive, it doesn't mean that it's not true. People differ in ability: some have valuable skills and others don't.