The universal unit of work is a decision
Discover more from Deliberate Internet
The universal unit of work is a decision
All knowledge work boils down to making decisions, big and small. When people abdicate them, companies die.
Break. Move. Build.
Most of the blue-collar work can be traced to breaking stuff apart, moving stuff, and putting stuff together. How does a bridge get built? Somebody removes rocks from the quarry, somebody moves them to the designated place, and somebody carefully arranges them so people can move other stuff on top of it. The same goes for a sandwich, desk, and space shuttle. Break. Move. Build.
Knowing how to do those things is knowledge work. It is tempting to use the same break-move-build model for information: Read a few papers and books, copy relevant information, and publish as a new insight. Break. Move. Build.
But again - you need to know how. What information to copy? What constitutes a good narrative? When to hit publish?
Software is making moving information easier and easier. Spreadsheets, word processors, and personal knowledge systems claim to take all the tedious away so you can focus on what is essential.
Powertools and robots are making, breaking, moving, and building stuff easier in the physical world. Progress is slower than in the digital space because working with atoms is harder than dealing with electrons, but every year new tools take on more labor, leaving you with decisions:
What tool to use and how to apply it?
Decision as the universal unit of work
All work can be traced to decisions:
As Michelangelo: Where to hit the marble stone with a chisel?
As a journalist: What information is relevant to my readers?
As a developer: How do I apply our design language to this particular Github issue?
As a manager: How to split this big task into manageable tasks?
As a creator: When to say “done“?
As a programmer: Should this be written in a functional or procedural manner? What library should I use?
As an author: What sources should I include?
As a security guard: Should I intervene? Should I call the police?
As a clerk: What shirt would fit this customer better?
Your workday is comprised of thousands of decisions ranging from very broad to very particular, but each one is a decision that needs to be made. This leads me to a conclusion:
Your job is about making decisions regardless if you are an Individual Contributor or a CEO.
In hierarchical organizations, we often expect the decisions to be already made by “people upstairs“, and to an extent, they need to provide a direction.
At each level of the organization, you make more and more impactful decisions. As an individual contributor, your decisions are mostly implementation details. Your boss may weigh in on the overall product direction, and his boss deals with the company strategy.
They pass on broad decisions to the level below. Individual contributors pass decisions onto the tools they use.
This fits with the concept of “Extreme Ownership“ greatly articulated by Jocko Willink:
Take responsibility for leading everyone in your world, subordinates and superiors alike.
If someone isn’t doing what you want or need them to do, look in the mirror first and determine what you can do to better enable this.
Don’t ask your leader what you should do, tell them what you are going to do.
Some downstream effects
Since all work is decisions, it cannot be accurately measured or estimated. (But you can and should measure outcomes)
Conversely, waterfall project management is less useful the better tools get because it assumes making decisions upfront.
Artisans deliberately mix blue-collar and white-collar work to tighten the feedback loop between the decision and the outcome (as they immediately see results). They put skin in the game to get better at decision-making.
The best way to be more productive is to make decisions faster.
The best way to slow down the entire organization is to make fewer decisions.
All management is passing more and more granular decisions down and asking and providing context up. Each layer of the organization is responsible for the appropriate decisions.
The best products make decisions for me, so I don’t have to. They are opinionated and have sensible defaults. This is because the companies behind them put in the work to make these decisions instead of abdicating them to their customers.
The urge to abdicate
Making decisions is hard. It takes brainpower, you can fail, and you can get punished. It’s scary. So it’s easy to convince yourself that it’s not your job to make them. Some strategies involve:
Abdicating decisions to superiors and waiting for their input
Abdicating decisions to tools/algorithms “Oh, the Algorythm™ will figure it out“. This may make sense to an extent, as some tools come with sensible defaults. But this is a decision you abdicated in favor of the tool maker.
Abdicating decisions to our AI overlords: “ChatGPT, please write me a business plan“.
Abdicating decisions to random events. I wrote about that strategy in “dont shoot yourself in the foot“.
Abdicating decisions to vendors/partners.
Abdicating decisions to your customers: “Let’s make this an option“.
Picking your battles is fine, and narrowing the scope of your work is probably a sane strategy.
But if you are a knowledge worker, your job is making decisions. Don’t abdicate all of them. Don’t become a tool.
Few things I’ve read
On Google’s challenges:
“The maze is in the mouse” is a fantastic rant on the organizational challenges facing Google today:
Like mice, they are trapped in a maze of approvals, launch processes, legal reviews, performance reviews, exec reviews, documents, meetings, bug reports, triage, OKRs, H1 plans followed by H2 plans, all-hands summits, and inevitable reorgs. The mice are regularly fed their “cheese” (promotions, bonuses, fancy food, fancier perks)
Most of the challenges boil down to abdication of decision-making:
While two of Google’s core values are “respect the user” and “respect the opportunity”, in practice the systems and processes are intentionally designed to “respect risk”
But I won’t work harder to prevent people from making mistakes — it is easier and more effective to work slower and slow them down. Just ask a few more clarifications and schedule another round of meetings two weeks from now.
Respect each other” is translated into “find a way to include and agree with every person’s opinion”. In an inclusive culture (good —it doesn’t withhold information and opportunity) with very distributed ownership (bad)
Death Spiral of Bullshit:
Austen (CEO of Lambda School Bloom Tech) put together a fantastic thread about why big organizations fail:
All the solutions to the death spiral of bullshit focus on preventing abdicating decisions:
* Teams as small as possible
* Extremely high talent bars
* Explicit decision-maker + disagree and commit
* Permission assumed in all things
* High tolerance for the (good) failure
* High tolerance for (good/non-infra) technical debt
* Zero tolerance for not being in details
Pretty soldiers
Presented without context:
