I Replaced 3 Roles in My Business With AI. Here's What I Learned.
You are using AI to work faster. Your competitor is using it to make that same work redundant.
There are two kinds of founders right now. Both are using AI. Only one is pulling ahead.
The first asks: Where can AI help with the work we already do?
The second asks: What work does AI make unnecessary or completely redundant?
Same tools. Two different questions. Leading to completely different business outcomes.
The first gets you faster output. The second gives you an unfair advantage. And the gap is widening every quarter.
Here is what this distinction looks like up close.
A founder I coach runs a $6M B2B SaaS company. He was nine months into trying to hire a head of content. Five candidates, three offers, two declines later, he was frustrated.
He kept blaming the market for great content people. He could not find one person who could think, write, and ship at his level.
His closest competitor was publishing two pieces a week. Every piece sounded like the founder, and every piece was getting read. They had no head of content.
“Have you ever written your thought process down on a document?”
He said no. It lived in his head. His competitor had built it — sixty pages, updated weekly.
When I asked him to describe the person he had been hiring for, he talked for five minutes. He listed his market thesis, his positioning, his customer conversations, his objections, his frameworks, and his sense of what good looked like. He had been searching for someone who already knew what he knew.
His competitor had built that document. An AI wrote every draft. Two senior people reviewed each piece.
He had been looking for the right person. They had been designing the right system.
Hiring excellence vs. designing it
You
9 months. 0 hires.
A job ad open on a laptop. Looking for someone who already knows what you know.
Competitor
60-page brief → AI → 2 reviewers
They wrote it down. AI drafts. Humans review. Two pieces a week, every week.
The brief is the hire.
That founder is not alone. I see this pattern across every industry I work in. Here are three signs to look for in your business.
The Three Patterns
Each one looks like smart adoption. None of them are.
AI adoption rates tell you nothing
A founder I coach told me his company had spent $5,000 on AI tools last quarter. I asked him what got cheaper, faster, or eliminated as a result. He could not name three things.
When founders tell me “eighty percent of our team uses AI weekly,” I ask the same question. What no longer happens because of it? Most cannot answer.
If your dashboard tracks who uses AI, you have a rollout. If it tracks what changed, you have a strategy.
Name three things AI has changed in your business. Faster does not count.
Drafting emails is the smallest use of AI
AI holds context, sees what you miss, runs analysis in seconds. Most founders use it to write things. The deeper use is in your hardest decisions, when nobody else has the data to push back.
I use AI before every major decision now. Before a hire, I ask it to argue against the offer. Before a pricing move, I ask it which customers lose. Before a launch, I ask it to write the strongest case for why this will fail. Half the time it shows me something I missed. The other half I move with more conviction.
Most use AI to type faster. A few use it to think more clearly when it matters.
Before your next big decision, spend twenty minutes letting AI argue against it. Ask for the strongest case against the move.
The work you defend hardest should not exist
Think about the role on your team you would never automate. The person whose work feels untouchable. That is the role this pattern is about.
Most defended work falls into three categories. Work somebody has been doing for years, where the system has built around them. Roles you built yourself, where replacing them feels like replacing a piece of you. Work where the conversation about changing it would hurt somebody you care about.
AI exposes this kind of work fastest because it does not see the history. It sees what the role does and how much of it could be a system.
You do not have to fire anyone. Rebuild the role around what AI cannot do — judgment, relationship, taste. Let AI handle the rest. The person stays. The company gets better.
Pick one role where 60% of the work could be done by AI today. Sit with this: am I protecting it because it is strategic, or because the conversation would hurt?
Your Blind Spot
Which pattern hit closest to home?
That is probably the one costing you the most right now.
Stop trying to hire excellence. Start designing it.
Fix Your Blind Spot
What role have you been hiring for that should be written down?
If you cannot answer that, you are still hiring for a role when it could be a system. A system that costs less, runs faster, and never stops working at 6pm.
Over the last three months I have replaced three roles in my own business with AI: my executive assistant, my content writer, and my entire web development team. Next on my list are my advertising team and my video editor.
In every case I started by writing down what the role actually did. Once that document existed, AI handled almost everything else.
Your turn.
Take one role you have open. Write what it actually does, the way you would teach a new hire on day one. Except this hire is a system. Then ask whether you need the hire, or you needed that document.
The AI Leverage Playbook
Find where AI is worth real money in your business.
A free 5-step audit for scaling founders. Built from coaching conversations and shipped systems — the same framework I use before automating a role, so you stop buying AI tools and start finding AI leverage.
- A 5-step audit to find where AI creates real ROI
- A live opportunity-cost calculator for any process
- A 30-day implementation plan you can run yourself
The Blind Spot
Every Tuesday. One pattern that is quietly costing you.
Most business newsletters tell you what to do next. This one shows you what is actually in the way. For scaling founders who are done chasing symptoms.
Browse all issues
The Blind Spot Archive →