The Solopreneur AI Stack: From $0 to First Paying Customer

If I had to start a solo business from zero today — no audience, no portfolio, no email list, $200 in the bank, 90 days — I wouldn't read another book on entrepreneurship. I'd run a tight playbook with AI doing the heavy lifting and human me doing the parts AI can't yet do.

This is that playbook. The tools, the prompts, the weekly cadence, and the mistakes to skip. It assumes you have one real skill someone might pay for. It doesn't assume an audience, money, or special talent at marketing.

The 90-Day Frame

Three milestones, one per month:

Month 1: Pick the niche, build the offer, set up the stack.
Month 2: Talk to 30 potential customers. Ship one free pilot.
Month 3: Convert the pilot into a paid client. Land #2.

If you hit those three milestones, by day 90 you have proof of concept and at least one paying customer. That's the win condition.

Notice what's NOT on the list: building a website, growing a Twitter following, writing a 12-week launch sequence, designing a logo. Those are all valid eventually. None of them generate revenue in the first 90 days.

Month 1, Week 1: Pick the Niche (5 hours total)

A "niche" is the intersection of three things:

A skill you have that someone with money would pay for
A specific kind of person who has that problem urgently
A reachable channel where you can find that person

Most solopreneurs get the first two right and skip the third. They pick "freelance writing for SaaS startups" — fine niche, but how do you reach a SaaS founder? "Cold LinkedIn" is not a real answer.

The AI prompt I'd use for niching:

I'm starting a solo service business with these constraints:
- Skill: [your skill]
- Time per week: [hours available]
- Money to invest: [$ amount]
- Channels I already have access to (people, communities, jobs):
  [list 5 channels]

Suggest 5 niches I could plausibly serve in the next 90 days.
For each, tell me:
1. Who specifically has this problem (1 sentence)
2. The single biggest reason they'd hire me over a competitor
3. Which of my channels would reach 5+ of them this week
4. The probable price ceiling for the service in this niche
5. The biggest reason this niche could fail

Run that prompt with Claude. Pick the niche where the answer to #3 is something you'd actually do.

Month 1, Week 2: Build the Offer (3 hours)

The offer is what you sell, exactly. Not "writing services." Not "consulting." A specific outcome with a specific price and a specific time-frame.

Bad offer: "I help SaaS startups with content."

Good offer: "I write a 4-post LinkedIn series for B2B SaaS founders that takes their best customer story and turns it into 4 high-engagement posts. $800. Delivered in 7 days."

The good version has: a specific buyer, a specific deliverable, a number, a price, a timeline. All five pieces. If your offer is missing any of those, you don't have an offer — you have a category.

Use Claude to stress-test it:

Here's my offer: [paste offer]

Pretend you're a skeptical buyer in this niche. Hit it with the
strongest 5 objections you'd raise before hiring me. Then for each
objection, suggest one tweak to the offer or one piece of proof
that would defuse it.

Iterate until the objections stop being scary.

Month 1, Week 3-4: Set Up the Stack (6 hours total)

Minimum viable stack to run on:

Claude Pro ($20/mo) — main work engine
Stripe (free until you take a payment) — payments
Calendly (free tier) — booking calls
Gmail (already have it) — outreach + delivery
A simple landing page — Carrd, Notion, or a one-page Next.js site
A Notion CRM — track every prospect, every conversation

Total spend: $20/mo. Don't add anything to this stack until you have a paying customer asking for it.

The landing page should answer one question: "Who are you and what do you sell?" Don't write three pages of about-us copy. Don't pick the perfect color palette. The page exists so a prospect can see you're real before booking a call.

Month 2: Talk to 30 People

This is the month most solopreneurs blow. They build a fancy funnel and wait for it to fill. It won't.

The job in month 2 is conversations. Specifically, 30 of them. Real conversations with real potential customers in your niche. Not "DMs back and forth" — actual 20-minute calls.

How to get 30 calls in 30 days:

10 from your existing network (former colleagues, friends in the niche)
10 from a community where the niche hangs out (Slack, Discord, subreddit, LinkedIn group)
10 from cold outreach using a personalized template

For each call, the goal is not to sell. The goal is to learn. Use this script:

"Thanks for hopping on. I'm trying to figure out [the problem your offer solves].
Could you walk me through the last time you dealt with this? What did you do,
what worked, what didn't?"

Then shut up and listen for 15 minutes.

After every call, note in your CRM: their exact words for the problem, their last failed attempt to solve it, what they currently spend on it (time + money), and whether they'd be open to a free pilot of your offer.

By call 30, you'll have so many patterns you'll be able to predict what call #31 will say. That's how you know you're done with research.

Month 2, End: Ship One Free Pilot

From the 30 conversations, pick the one prospect who:

Has the problem urgently (not "yeah, that bothers me sometimes")
Has the budget to pay if they liked the result
Will give you a referral or a public testimonial in exchange

Offer them the pilot for free. Deliver in 7 days. Over-deliver by 30%. Get the testimonial in writing the day after delivery, while it's still fresh.

Month 3: Convert + Land #2

Now you have one piece of proof. The pilot client either becomes paying client #1 (60% of the time, in my experience) or refers you to someone who does (30%). The remaining 10% are people who said yes for free but won't pay; they were never the right fit.

For the next prospect, you have a story:

"I just helped [pilot client] [specific result]. Could I do the same for you for $[your price]?"

That sentence — with a real name, a real result, and a real price — converts at 5-15% in cold outreach and 30%+ from referrals. It's the difference between "trust me, I'm new" and "here's what I just did for someone like you."

What I Wouldn't Do in 90 Days

The tempting work that doesn't move the needle:

A perfect website. A one-page Carrd or Notion page is enough.
An email newsletter. Audiences compound; you don't have time to wait.
A digital product. Service revenue is faster, higher per unit, and teaches you what people actually want.
A logo. Use a Google Font and your name in 36pt. Done.
Most of social media. Pick one platform where your niche actually hangs out, post 2x/week, ignore the rest.

The 90-day game is conversations and proof, not infrastructure.

What Comes After

Once you have 2-3 paying clients, the next 90 days are about systematizing. That's when AI starts paying real dividends — automating the parts of delivery that don't need you, building Claude Code skills for the recurring tasks, packaging your service into a productized offer.

But that's day 91 work. Day 1-90 is just: pick a niche you can actually reach, build a real offer, talk to 30 people, ship one pilot, convert it.

AI doesn't replace the discipline of doing the work. It just makes the work go faster once you're doing it.