The Wayfinder
The Automation Trap
The Trade Desk
Automating Chaos
There’s a pattern I see in almost every business I consult with. The owner comes to me and says: “I need to automate my processes.” What they mean is: “Everything is on fire and I want a robot to hold the hose.”
Technology amplifies whatever you feed it — including your problems.
But here’s the uncomfortable truth: if your process is broken, automating it doesn’t fix it. It just breaks it faster, at scale, with no one watching.
The Three Stages of System Failure
Every business that tries to automate before they’re ready goes through the same stages:
- The Honeymoon — “This is incredible, we just saved 10 hours a week”
- The Cracks — “Why is the system sending duplicate emails to clients?”
- The Blame — “The software doesn’t work” (it works exactly as instructed)
The software always works. The question is whether you told it to do the right thing.
“A computer lets you make more mistakes faster than any other invention in human history — with the possible exception of handguns and tequila.”
— Mitch Ratcliffe
Manual First, Always
Before you automate anything, do it manually 50 times. I’m serious. Fifty times.
Map it before you automate it. Every. Single. Time.
By the tenth time, you’ll notice the steps you always skip. By the thirtieth, you’ll have simplified the process by half. By the fiftieth, you’ll know exactly what to automate — and more importantly, what to leave human.
The rule is simple:
- Automate decisions that are repetitive and rule-based
- Assist decisions that are complex but patterned
- Leave human decisions that require empathy, judgment, or creativity
The AI Agent Myth
Everyone wants an AI agent. Few people need one — yet.
The conversation usually goes like this:
“I want an AI that handles my customer support.”
“Great. What’s your current support process?”
“We don’t really have one.”
That’s not an AI problem. That’s a systems problem. And no amount of GPT-4 is going to fix a process that doesn’t exist.
What AI Agents Actually Need
For an AI agent to work, you need:
- Clear inputs — structured data, not chaos
- Defined rules — decision trees the agent can follow
- Feedback loops — a way to catch and correct mistakes
- Human oversight — someone watching the machine watch the process
The best automation looks simple. That simplicity took months of manual work to earn.
" Don't automate your confusion. Systematize first. Automate second. Scale third. "
The Workshop
The Process Audit Checklist
Before you touch a single piece of automation software, run every process through this audit.
Step 1: Document the Current State
Write down every step, exactly as it happens today. Not how it should happen — how it does happen. Include the workarounds, the exceptions, the “oh, and sometimes we also…” steps.
Step 2: Identify the Bottlenecks
PROCESS AUDIT TEMPLATE
======================
Process Name: _______________
Owner: _______________
Frequency: _______________
Steps (as they actually happen):
1. _______________
2. _______________
3. _______________
Bottlenecks (where things slow down):
→ _______________
Failure Points (where things go wrong):
→ _______________
Manual Steps That Could Be Automated:
→ _______________
Manual Steps That SHOULD Stay Manual:
→ _______________
Step 3: Simplify Before You Automate
For every step in your process, ask:
- Is this step necessary? If not, remove it.
- Can two steps be combined? If yes, combine them.
- Does this step require human judgment? If yes, keep it human.
- Is this step the same every time? If yes, it’s a candidate for automation.
The best automation starts with the best documentation.
The Sanctuary
The Wisdom to Know the Difference
There’s a reason the Serenity Prayer has endured for nearly a century. It captures something fundamentally true about the human condition — and about building systems.
" The beginning of wisdom is this: Get wisdom, and whatever you get, get insight. "
Wisdom isn’t knowledge. Knowledge is knowing how to build an automation. Wisdom is knowing whether you should.
I’ve built systems that worked perfectly and served no one. Technically flawless, spiritually empty. They optimized for speed when the real need was for presence. They automated communication when what the customer needed was to be heard.
Efficiency vs. Faithfulness
The world tells you to optimize everything. Reduce friction. Eliminate waste. Move faster.
But some things are meant to be slow.
Not everything that can be measured matters. Not everything that matters can be measured.
A handwritten thank-you note takes ten times longer than an automated email. But it communicates something an email never can: you mattered enough for me to stop and write this.
“Whatever you do, work at it with all your heart, as working for the Lord, not for human masters.”
— Colossians 3:23
The question isn’t “can I automate this?” The question is “should I?”
The Archive
Resources for Thoughtful Automation
Automate the Boring Stuff
Al Sweigart’s practical guide to automating repetitive tasks with Python. The title says it all — automate the boring stuff, keep the interesting stuff human.
Read free onlineThe E-Myth Revisited
Michael Gerber’s classic on why most small businesses fail — and how systems thinking (not just hard work) is the answer.
Learn moreShape Up
Basecamp’s framework for building software in cycles. A masterclass in scoping work and shipping on time without burning out.
Read freeZapier Blog
Practical automation guides and workflows for non-technical operators. Good for learning what’s possible before you build custom.
ExploreFurther Reading
- Thinking in Systems by Donella Meadows → — The foundational text on systems thinking
- The Goal by Eliyahu Goldratt → — Constraint theory applied to business, told as a novel
- No Silver Bullet by Fred Brooks → — Why there’s no single technology that solves everything