Cut Before You Add
A lot of the CEOs I coach have confessed the same fantasy this year: they wish they could kill what they have and start over with just AI and three people. These aren't failing companies. They're healthy businesses run by people who spent years building them.
I get why, and it's not really about the technology. The fantasy is about the mess: years of processes nobody has questioned, tools stacked on tools, work that exists because it has always existed. What they want isn't agents. It's simplicity.
Then the call ends, and most of them go back and do the opposite. They add another tool, another automation, another layer, because the pressure to optimize treats every problem like a shortage of activity. But adding is how the mess got here in the first place. When more is your operating principle, more is what you end up with. In the sprint toward more, you end up with more work.
I built an AI-first company a decade ago, before plug-and-play tools existed. Five of my 26 employees weren't human, and we ran 96% margins. Even now, the second I catch myself doing a task twice, my first instinct is "how do I automate this?"
But at every step I ask one question first: what am I improving, and why?
That question keeps every tool and system serving you instead of the other way around. Without an answer, you're optimizing for the sake of optimizing. That has never grown a company.
When I started out, the question wasn't a choice. I was building with almost nothing. I couldn't pay for help and I couldn't do everything myself, so everything I built had to save me time or money. I had to build the machine so I wouldn't become the machine.
That constraint still drives me. The goal was never more. It's always less: less effort, less mess, less of the work that was never yours.
The discipline is reversing the core question. Not "where can I add?" but "where can I cut?"
A founder I coach came to a recent session ready to build. He wanted more content, more programs, a bigger list, new offerings in new places. It all sounded like growth. So I asked what all of it would get him.
He didn't have an answer. "Expand" feels like a direction, but it's fuzzy. You can post twice and call it expanding. We worked it down to a single number he wants his audience to reach in twelve months. The moment the number existed, half of what he was about to build fell away. It never served the number. It was just more.
That's the first cut: name the actual job before you build anything. The second cut is the one that kill-it-all fantasy is begging for. You don't need to burn the company down to get the simplicity you're dreaming about. You need to untangle before you automate. Automating a tangled process doesn't fix it. It pours concrete over the tangle.
So before you build anything on top of your operation, walk through it with three questions, in order. What can I cut entirely, because nothing depends on it? What can I simplify, because it grew complicated on its own? Only then: what can I automate?
And when you reach that third question, the answer is usually not a new tool. Most of a business can be automated by connecting two systems you already use. In my business today, when a client's invoice goes out, they're added to my newsletter list automatically. When my personal calendar gets an entry, my business calendar sees it, so nothing gets double-booked. Nothing new was built. Two things that already existed started talking.
When I built the sales system at Butler, I automated 70% of the pipeline with a scraper, Zapier, Mailshake, and email templates. It took a few days, with no engineering team. I broke down the layer-by-layer version here.
Cutting first is also what buys you the capacity for the tools that deserve real investment. Because when I do invest in a tool, I invest all the way. A useful AI agent takes the same work as onboarding a person. You teach it how you make decisions, what your standards are, what you believe about how a business gets built, how your operation actually runs.
You can see all of that in how I hand Claude any real piece of work. Every project opens with the same brief, and it's the same brief I gave every person I ever hired:
What you're doing. The whole job, spelled out.
Why you're doing it. The context that lets it make good calls without me.
What success looks like. So we agree on done before anything starts.
How I want the output. The format, the files, where it all lands.
Where you need my approval. Usually: nowhere. Act. Don't ask me.
Evidence to bring back. Every claim cross-checked against the sources, so nothing gets made up.
Then I point it at where everything lives, and it goes.
There's no prompt trick in there. It's delegation. It's the briefing a new team member needs to do the job the way you would do it. And just like a team member, over time you give less. It learns how you decide, and the brief gets shorter.
The brief also carries your beliefs. Here's a piece of my own setup: I don't believe in working hard. So any plan Claude gives me has to assume ease is possible. I want the shortest path, not the most impressive one. I want everything tracked and organized. I want it to get to the point instead of handing me a novel.
That work forces a clarity most founders never sit down for. If you can't describe what you do, no tool can do it for you. So you have to figure out what you do, in detail, out loud. And that clarity outlives the tool. In two or three years Claude will be replaced by whatever comes next, arriving with the same urgency and the same pressure to add before you know what you're solving.
The foundation of a good business will be what it has always been: simple operations. Less, not more. Cut before you add. Simplify before you automate.
Reply or comment and tell me about the tool or automation you've been building that you suspect you don't need. I read every one.
-Christine
If You Want To Go Deeper...
Here's where to start:
⚫️ Private Coaching: for early and growth-stage entrepreneurs who want to lead with more clarity while increasing their resiliency. See if we're a good fit here.
⚫️ The 20 Hour CEO Self-Paced Course:The frameworks, systems, and playbooks to stop being the bottleneck — on your schedule.
⚫️ The 20 Hour CEO Live Cohort: 3 weeks, 6 live sessions. Bring your actual business, rebuild how it runs, and leave with systems already in motion.
