Why I Told a Friend to Stop Building an AI App
How 3 people I work with make $6K/Mo with Claude. No app. No quit.
Two weeks ago, a friend texted me.
He’d been reading about AI businesses for months. Watching the YouTube creators. Lurking in Discord. Buying the courses. He wanted to know where to start.
I wrote back two words.
Sell a service.
He told me that wasn’t what anyone else was saying.
Everyone else was telling him to build an app.
So let me explain why I’m not.
The lottery you don’t have time for
For every AI app that makes a hundred dollars, there are roughly ten thousand attempts that never make a dollar.
Even if you can engineer, the build is barely half the work. The other half is making something people will pay for. Marketing it. Finding customers. Supporting them. Fixing what breaks.
That is lottery odds.
And the lottery takes months of your evenings before you find out if you drew a winning ticket. If you already have a real life in motion (a day job, a family, bills that depend on income actually showing up), you do not have that runway to burn.
So you flip the equation.
The service play
A service flips the risk.
Build an app: You spend → you ship → you pray.
Sell a service: They pay → you build → you ship.
Someone pays you first. You deliver after. You are selling a promise, and you build the specific deliverable with their money.
And the buyer doesn’t care if AI is doing the work. They care about the outcome.
Nobody asks a lawyer which software they used to draft the contract. They ask if it protects them.
The service business runs on exactly that logic.
Three people, no app, up to $6K a month
I work with three people who built this kind of business on the side of their day jobs in under ninety days.
Maya was a senior PM at a SaaS company. She writes competitive intelligence reports for startup founders. Twenty-five hundred per report. Five thousand a month by week eight.
David did eight years in management consulting. He writes investment memos on nights and weekends. Three thousand per memo. Six thousand a month by week twelve.
Priya did twelve years in B2B marketing. She runs a content strategy retainer service now. Thirty-five hundred a month by month three.
None of them built an app. None of them learned to code.
They sold a service.
They let Claude do the work.
The four-step system
I call it The Build System. It’s how you turn expertise into a service business without giving up your day job.
1. What to sell. → A service, not a product.
2. How to find an idea. → Where your old expertise meets AI compression.
3. How to deliver. → Build a Claude Skill once. Reuse it forever.
4. How to sign clients. → Ten real conversations with the right ten people.
Plus a 30-day plan you can start this weekend.
That is the playbook.
The part where most people stall is step 2.
The part everyone freezes on
Two questions:
What have you been paid to do at a senior level for more than three years?
Where does AI now make that work at least ten times faster?
The overlap between those two answers is your service.
That overlap is what I look for when I work with someone one-on-one. It is also what took me the longest to systematize across different industries, different career levels, different constraints.
So I compressed it into a Claude Skill.
It is called the Second Income Map.
You load it into Claude. You answer seven questions about your professional history. It produces a one-page service brief in about fifteen minutes:
Your specific service direction
Your target customer
Your price band, anchored to pre-AI market rates
A five-conversation validation test you can run this weekend
The single next action
I am giving it away free.
Download the Second Income Map skill here.
Install it in Claude. Run it once. See what overlap your career hands you.
If the answer feels obvious in retrospect, that is the point.
— Guney
P.S. If a friend or coworker has been asking you the same question my friend asked me, forward them this email. The Map is free.