[ssivark]

The real challenge is attention management. Switching gears from an execution orientation (knocking off tasks) to a careful thinking orientation is hard. The hardest aspect of deep work (for people used to a structured form of productivity) is immersion in the task/topic without a concrete plan of "how" to tackle it. (The necessary immersion is what makes "deep" a good metaphor) Every such session is a risky investment where you don't know what you'll come up with. Unless you believe it's a risk worth taking, trying too hard to de-risk endeavors will rob you of opportunities for deep work.

Deep work involves situated thinking -- thinking while doing, making use of feedback.

OTOH, the structured approach biases towards separating the two closely knit parts into "planning" and "action" -- so much so that in our organizations the two are done by different people/teams!

Drucker said that a decision has been made only when it's clear what the actions are. Correspondingly, deep work is required to the extent that the analysis can't be front loaded in that manner.

[sound1]

Beautifully articulated! On a tangent, I always seem to have issues with my superiors because of the way I work. Most of the initial part of a project I do nothing but think about how to approach it, best ways to solve some of the problems and what pitfalls to avoid. During this period they don't get much of the status updates or progress and they get mad. Then in a short burst I finish off everything when they think that project may be delayed, and this gets them extra mad because according to them I have been secretly working hard but never told them. I really don't blame them but sadly I don't work like most people do and admittedly it is a bit of a problem for project managers..

[jogjayr]

It may help to document ideas and pitfalls as you think of them. If you're trying out things in code, leave that work on your own branches, and link from the doc. It gives managers a tangible artifact to see you're making progress. It's also valuable documentation for others after the task is done. Extra plus: come performance review time, you have a paper trail of everything you've done, everything you tried that didn't work out, everything you can take credit for.

[Denzel]

Yes, this is fantastic advice. The most effective senior engineers on my team almost always create what’s called a “start doc”. Written at a high level, the doc discusses background context, enumerates possible solutions with a pro/con list for each, and records open questions and out-of-scope issues. They’re usually only 1-3 page gdocs.

The most important part is that it opens up discussion to the team. Everyone is able to read, comment, and offer up improvements or get clarity on questions they have. After a few rounds a feedback and revision — usually 2-3 days — a decision is made and the engineer goes off to break out a few tickets and start the work.

It’s async, collaborative, efficient, visible, and just a generally pleasurable process as both an individual and teammate.

[sound1]

Great advice, thank you!

[snlacks]

There's no reason why "figure out what to do next" can't be a concrete task. I set up a daily schedule and I schedule reassessing and specifying more detail into my work.

[abhiyerra]

I set aside a day of the week, Tuesday, for all my planning for that week. Basically the deep work for that day is figuring out the plan for the week. The rest of the week is accomplishing that work. Basically Scrum applies to deep work.

[who8mylunch]

Thank you for this concise summary. At my job I'm often accused of not being a team player because I refuse to get sucked into other people's shallow-work emergencies. It feels good to know that I'm not alone!

[commandlinefan]

> The real challenge is attention management

See, I think you have that backwards - the problem is management attention. When management is giving you their full attention and micromanaging, meaningful work becomes impossible.

[dc443]

I like what you said but i think grandparent's use of management is not the management that you're referring to.

[commandlinefan]

I was going for wordplay but I guess I went too obscure.