5 ms·
The "Overkill Cult" is an artifact of a working world built on subordination rather than progress. When the VP is working till 7:00, the Director better be ther
by michaelochurch 12y ago
The "Overkill Cult" is an artifact of a working world built on subordination rather than progress. When the VP is working till 7:00, the Director better be there with him and then some (if the VP delegates out the door). In many companies, this ripples down the chain. It's a very old rule: don't sit when the king stands, and never appear more relaxed than your boss. To not participate in shared suffering is taken as rude, even if there's no good reason for that suffering to exist. Unfortunately, an attempt to remove that suffering from existence will upstage the VP and end one's career even faster.
The Big Lie is that most people aren't being paid for their intelligence or creativity or even their "potential", but for their availability. They're excess capacity. For a variety of social reasons, though, management needs to come up with 8 hours per day of work for them. Worse yet, there's a fucked-up incentive structure. Since a manager's job is often to make work, set meetings, and change what people are working on, you have a culture where overachievers end up overmanaging and that usually devolves into micromanagement or some other pathology. If a manager over-does his job, you don't get an efficient organization: you get monstrosities that look like Scrum. A good manager over-does his job upfront and very briefly (to remove standing problems) but then sits back and underworks at the management thing. His job is to make himself not needed.
Before about 1975, the culture of subordination made sense. It's an age-old pattern: subordinates do the work, and dominant people "lead" or "manage" or "order". It's failing us now because computers simply outperform us in all ways-- availability, cost, reliability, context-agnosticism-- when it comes to subordinate work. They work 24-hour days and don't give a shit about whether they're being trained up to something better (like their boss's job). We now live in an era where there's a strong social pressure to work in a way that makes one eventually replaceable by a machine or, at least, a person of lower skill (e.g. a "commodity Java programmer" instead of a hacker magician).
The old industrial regime didn't only require widespread subordination, but it required layers of educated, intermediate subordinates. They had to be given a sense of importance (19th century clerks, 20th century middle managers and "junior executives") but their working lives needed to be mediocre, because making people into a machine is hard and making a large group of people into a machine is so hard that it demands hierarchy and it demands an elevated-but-still-subordinate tier. This need for intermediate subordinates meant that our (errant?) tendency, in human organizations, to promote people based on their being good subordinates, rather than creative ability or leadership, didn't hurt us that much-- not back then. Now it does, because this skill of turning a mass of people into a machine is far less useful, because actual computing machines are better machines than we could ever be.
Unfortunately, I don't see this changing, because humans are short-sighted and too often focused on relative dominance rather than absolute prosperity. It's like the Chinese finger trap. Technological unemployment and underemployment are real and the severe mismanagement of the prosperity that technology generates means that average people (and even computer programmers) have to fight harder to get the dwindling number of jobs. You don't deliver genuine progress by "fighting"-- it takes time and reflection and collaboration-- but our instinct amid scarcity is to fight each other and the behavior of competing-to-subordiante (as in a chieftain's harem, or in capital's harem, also known as the corporation) is quite old as well. You get the macho-subordinate culture (i.e. a race to the bottom) of Scrum and open plan offices (violent transparency). It's a losing battle because we're fighting to do something that we, as humans, are no longer competitive at. Ultimately, anything that can be done by a ScrumDrone in 2015 will be done by a compiler or a tool in 2030.
Personally, I would rather accept this and focus on progress instead, but it's hard to make that statement without drawing the "you just don't want to be a subordinate" rebuttal (and that's true, I don't) and be painted as "not a team player". So little progress is made and, from what progress does occur, the prosperity it generates is badly invested. Thus, we have a world in which people work harder and more productively, but for a personal take that is stagnant at best and quite possibly (considering hyperinflation in the Satanic Trinity of housing, health expenses, and tuition costs) declining.
Long hours persist because, ultimately, the most measurable way to subordinate is to spend an ungodly amount of time in the office, which is a place where no one-- especially not the people who enjoy the work and want to be productive-- really wants to be.
- wavefunction 12y agoI always enjoy your posts and insights. I find them to be right on when I consider my own experiences at various companies (all outside the mythical and magical Vale of Sillicone).
- mmanulis 12y agoWhat role do you see managers playing in an organization? What do you mean by "His job is to make himself not needed"? It seems like for any organization, some structure is required just for information dissemination. So making an enormous assumption that corporate leadership trusts their people to do the job and do it well, what does a structure that is not a "giant human machine" look like? Thank you
- bkeroack 12y ago(not the OP) The job of a manager is to lead as a natural outgrowth of deep domain knowledge and industry experience. The best bosses I've ever had were those from whom I learned new things every day, and I was glad to take guidance and instruction from them. The trust goes both ways--the manager trusts in his/her people to make the right decisions because he/she has led them well. In that way they don't need to be involved in the day to day of their reports and are only called upon in moments of uncertainty or conflict. On the other hand, I've had bosses where I was 95% certain I could do their job better than they could, and the only reason I had to subordinate to them was because they got there first. These were also the type of people to micromanage to 'prove' that they had organizational value (probably also because they feared losing any control). You then have a few choices: either shut up and submit, or play the political game of positioning yourself either as a lackey of the boss or as a potential replacement should they fail (there's obviously some incentive there to secretly engineer situations to make them look bad/incompetent to the rest of the org). Neither is a particularly good option--that's when I know it's time to seek better opportunities elsewhere.
- mmanulis 12y agoI'm not convinced that leading is the job of a manager, at least not all managers. As an example, a line manager, someone who manages the work of several developers, is more of a Project Manager / People Manager and less a leader. Maybe this is getting too much into semantics. A leader, as I tend to think of that position, is about setting direction, guiding the vision, dealing with the culture, more strategic in whatever they are leading. This could be a company or a project or a product, doesn't have to be high-falutin, can be as simple as leading a project to completion. The focus, however, is really about the strategic view of whatever it is that's being led. A manager is the tactical view. Making sure people have enough work and that work fits the project. Making sure that the people doing the work are the right ones to do that work (some crossover into leadership here, can't help that). A manager is really part communication traffic cop (in a positive sense, not "give you a ticket" sense), part nanny (reward & punish behavior), part mentor and part canary for the upper management of things going wrong. As a side note, this is why I found 1-1s valuable at all levels of managing. When I was running several teams, making sure my leads were having 1-1s with their people and talking about how their teams are doing gave me a view I would not have had otherwise. I'm still learning about this though. Part of my thinking about the original comment was about the role of management in an ideal scenario. If you remove the "human machine" from the conversation and look at what most of the time managers at all levels are doing, it's communication management. Yes, this is an overgeneralization to a great degree, but it helps me to think about management in that way. If your manager built his/her interactions with you around things you need to know (communicating vision, direction, business needs, etc.) from above and took your feedback and communicated it up, then you get lots of conversations and feedback loops. This requires the upper management to listen to what the people on the ground are saying, which is rare in my limited experience. Incidentally, what you get is managing the what and not the how, letting individual contributors deal with the how. Which opens up a whole can of worms for corporations and, sometimes, from a legal perspective too - think SOX & HIPAA as examples. Bottom line, I believe management is to leadership as tactical view is to strategic view. Genuine communication is what management is all about, makes our work interactions human and makes it easier to give and receive feedback across the entire org. Of course I just described a utopia, but it's nice to work towards that, even if it's just a person approach.