Great Systems Scale Faster (Take The Time To Build Them)

[ben]

After thinking about yesterday’s Q&A about @chris_becker ‘s AI Won’t Save You’re Thinking article, it reminded me something super important, and a pattern that keeps popping up in my life that everyone wanting to perform at a high level in the product realm should know.

Great systems scale, immensly. No, not just 2-3x your work. Not even 10x your work. With AI, you get outputs that are now easily 100x the quality and quantity.

Why? Everything compounds. You, having a faulty understanding of your own product, your own business, the users that are or could be using the tool, are extremely limiting you. Not just holding you back from these 100x gains. You’re rescinding backwards.

AI creates two pathways for you: you’re either increasing this comprehension gap, further detaching yourself from reality just by doing outputs, or you’re extremely grounded, pushing further into the right areas because you’ve done the work, and taken the time to understand what the hell is going on.

Build Great Systems

Losers have goals. Winners have systems. - Scott Adams

I’ve been reading more and more of Scott Adam’s work, specifically How to Fail at Almost Anything and Still Win Big (no, I’m not sponsored). Actually, across a lot of my recent reads, I’ve been seeing this trend around systems happen more and more. I’m also seeing the benefits of implementing systems over just having a plain-old goal.

The core point is: Systems compound. Goals do not. You can iterate on a system. Build it bigger, remove things that don’t work, and keep that process for infinity- as the times change, as personal needs change, pretty much whatever. You can also easily kill the systems entirely (as long as you understand the cost-sunk fallacy).

Here are the three key principles for building great systems:

1. Holistically. Understand the requirements, needs, side-effects looking from a larger frame.
2. Repeatability. You need to be able to repeat it, frequently. That way even on bad days you can “make it happen”.
3. Peele the bandaid. Remove solutions that are temporary. You need long term fixes to solve the underlying problems.

Having all of these help create great systems, can extend to pretty much anything you can think of.

Here’s a great video that sums up building systems that work:

How to Build Systems to Actually Achieve Your Goals

A few questions that I do have to us system builders out there:

• What are the principles that you think are important or that you follow for building out successful systems?
• What are some pre-built systems that you currently follow to either help with your workflows, teams, or even personal life?
• Who’s a great resource for either providing system templates or helping others build them for custom requirements?
• How do you use AI to further extend system capabilities?
• What AI systems have you created to extend the quality of your own work?

[nathaliesmith]

@ben

The core point is: Systems compound. Goals do not. You can iterate on a system. Build it bigger, remove things that don’t work, and keep that process for infinity- as the times change, as personal needs change, pretty much whatever. You can also easily kill the systems entirely (as long as you understand the cost-sunk fallacy).

Soooooo, yes. But also I’d argue that systems compound, only in the direction they’re pointed at. Seems like without a clear goal or objective, you can get really efficient at the wrong thing.

[ben]

Agreed! It’s not that goals are useless, but people do forget about the system part pretty frequently.

You need BOTH.

[nathaliesmith]

[Bryan]

@ben

I’ve been reading more and more of Scott Adam’s work, specifically How to Fail at Almost Anything and Still Win Big

Good read.

Things I’ve learned in systems, and things I’ve been wrapping my head around lately:

Start with the decisions
If a system doesn’t clarify decisions that need to be made, it becomes overhead. Every system I build has to answer: What decision does this help us make faster or better?

Build systems around signals
Process can mask activity for activity’s sake. I care about signal. What are we learning? What changed? What do we do next? Feedback loops are gold.

Visible systems are easier create momentum
If only one person understands the system, it is fragile. The best systems create shared language. They give teams a common way to talk about impact.

Keep it lightweight at the edge
The closer you are to execution, the simpler the system should be. Heavy systems slow down iteration. That’s not to say larger systems don’t have complexity to manage.

Pressure test under constraints
A system that only works when you have time, money, and alignment is not a system. It is a fantasy. You need the pressures of constraints.

[ben]

@Bryan

What decision does this help us make faster or better?

Love this!

@Bryan

Visible systems are easier create momentum
If only one person understands the system, it is fragile. The best systems create shared language. They give teams a common way to talk about impact.

This insight is super cool. Are there any particular ways to lean into this?

Great stuff @Bryan !

[nathaliesmith]

@Bryan

Start with the decisions
If a system doesn’t clarify decisions that need to be made, it becomes overhead. Every system I build has to answer: What decision does this help us make faster or better?

Loooove this one – 10/10