They best managers I have had are stealthy. They'll monitor from afar mostly (while getting their work done). They get updates during appropriate times. They respect their direct reports and understand that one can't get anything done with constant interruptions. They shield their direct reports from outside interruptions as well. They do, however, interrupt when they get a sense of something slowing down or blocking. They'll work with the individual or team to quickly get things moving again.
Sadly, the former outnumbers the latter by like 9:1.
One advantages of being the "little independent consultant" is that I can control my environment. I think you can reduce a geniuses output by a fair percentage just by adding distractions (like all sitting at the same table as your colleagues).
Same here. It’s actually kind of mind-blowing how little I get done at work compared to hobby projects, despite feeling as though I’m working much harder. In my free time, I’ve written my own global illumination software from the ground up, my own website, and a few open source libraries. In the same time at work, maybe one tiny feature for one internal tool. I understand why it happens and why it is necessary to some extent, but it’s much less satisfying than seeing the substantial day-to-day progress that can be accomplished on a solo project.
You probably think more before you write. The time away (at the very least the time while spent on other work) gives you time to figure out what to do next, whether it's passively or actively. You're more prepared when you finally do something.
You can categorize your interruptions to see where your org is failing. For me, it's people dropping in to ask questions because it is easier to walk by and get an answer than to dig in to the problem and find it themselves.
Or to phrase it in a slightly bitter manner, the people in charge are mere humans like us with limited field of view, and the fact that they think they can capture everything in an MBA paradigm is a mismatch which generates avoidable inefficiency.
This last weekend I did something that I normally refuse to do, which was to catch up on work. I still dedicated most of my time towards non-work, yet being able to work without interruption was not only a breath of fresh air but seemed to allow me to get more done in the same amount of time.
My guess as to why organizations don't care about focus is that they value the perception of "getting things done"; when engineers that need focus are frequently interrupted by the process, they'll shift to getting smaller tasks done first so that management will know that they're working on something. Organizations will tend to tolerate a more complex task taking longer than estimated, but in my experience are less tolerant of engineers not pushing lines of code for days because they're taking time to implement said complex task in a good way. In essence, it's a spin on "butts in seats", except it creates the illusion of productivity when in reality it could mean counter-productivity.
I'm not exactly sure that having the process wall people off from interrupting each other is necessarily a good thing. What would be better is a good planning process that limits unnecessary change combined with a working culture that values focus.
EDIT: I'm not talking about legitimate interruptions like when someone needs help or wants to pair program. I'm talking about interruptions that are either environment related or are a result of process failure(e.g. new tasks and course corrections frequently added to the pile because of poor planning). Although I can imagine some would want to limit "good interruptions" to a window, if possible.