You don't need money, a product, or an audience to make your first dollar online.
You need one useful skill and one person with a problem.
That's the whole pitch behind Brett Malinowski's 71-minute course on how to start an AI services business from absolute zero. And Brett isn't theorizing. He sold video services in college, made six figures his first year after graduating, then ran a dev agency called WGMY Labs for 3+ years charging $6,000/month for unlimited software development. That agency made him over $1M a year.
I watched the full thing so you don't have to. Here's the playbook, with the numbers.
Why services beat products when you're starting from zero
Every SaaS founder I interview on the podcast had to solve the same cold-start problem: months of building before the first dollar shows up.
Services skip the wait entirely.
You find somebody with a problem. You offer to solve it for money. There's nothing to build before you're allowed to earn.
Brett's proof points from the video:
- He made six figures in year one out of college selling video and paid ads services.
- WGMY Labs charged $6,000/month per client for unlimited dev work and cleared $1M+/year (and sometimes took equity in the apps they built).
- Evan, a 20-year-old journalism major, made $17,000 in his first 3 months selling clipping services. While still in college.
That last one isn't a fluke. It's the same Evan whose clipping agency later hit $35K/month. I broke that story down in how Evan built Clipping Culture from $88 to $35K/month.
And a service business doubles as the best R&D lab for a future SaaS. You get paid to learn exactly what a niche will pay for. Brett's agency built apps for creators, took equity, and learned the whole market from inside client work.
The 7 services Brett says are in demand right now
Pick one. Not three. One.
- AI implementation. Helping businesses install and actually use AI in their operations. Every business knows AI matters, almost none know where to start. That gap is the service.
- UGC (user-generated content). Short-form videos for brands, filmed on your phone. Influencers got expensive, so brands shifted budgets to everyday creators. Fastest path to a first dollar.
- Website building. Every business needs one, and half of them are outdated or ugly. With AI tools the build takes hours, not weeks.
- App building / software development. Brett's own path. Software is never finished, so it naturally becomes recurring revenue. Plus you learn to build your own SaaS on someone else's dime.
- AI video creation. Veo 3, Kling, and friends now do what used to take a production team days.
- Paid ads. Running Meta, TikTok, and Google campaigns for businesses. This is what Brett did in college.
- Clipping. Turning podcasts and streams into shorts. Around $1 per 1,000 views on content rewards platforms, and a clipping agency can charge 30% of a client's campaign budget (that's $3,000 on a $10K campaign).
If you want the deeper argument for why the services model beats building software first, I covered Brett's earlier video on that in why the most profitable AI business isn't software.
Positioning: the 401(k) story that shows how niching actually works
Best part of the course, by far.
One of Brett's students started with "I help businesses implement AI." Nobody bit. Too vague, no advantage.
He was a finance major, so he pivoted to financial advisors. He understood their world and how they made money. Then, on sales calls, he noticed they didn't care about "AI implementation" at all. They cared about one thing: getting more clients.
So the offer became "I help financial advisors get 5 leads per month with AI outreach."
Then it got sharper. He found a subset of advisors who specialize in 401(k) transfers for people leaving corporate jobs. His final system: an AI that tracks job changes on LinkedIn, and when someone at a 250+ employee company changes positions, it automatically messages them from his client's account offering help with their 401(k) transfer.
A very specific group. A very specific result. A very specific method.
→ That's why he charges premium prices while generic "AI consultants" fight over scraps.
Your positioning exercise is the same: start with a skill, pick an industry you understand, then let real conversations tell you what people will actually pay for.
Getting clients: 25 leads a day, on autopilot
Brett's client acquisition system is embarrassingly simple.
Step one is a lead tracking sheet: name, business, platform, contact info, status (not contacted, messaged, replied, booked, closed). That sheet turns "I should do outreach" into a sales pipeline you can actually see.
Step two is automating it. Brett uses Claude's desktop app plus its Chrome extension so the AI can browse, prospect, and fill the sheet itself. His setup: a scheduled task fires at 9:00 a.m. and finds 25 new leads, then at 10:00 a.m. the AI starts messaging them with his pitch.
25 personalized touches a day, every day, without him lifting a finger. That's 750 prospects a month while your competitor sends 10 DMs and gives up.
This kind of system isn't just for solo operators either. Agencies like TwiLead build business automations like this for established companies. If businesses pay agencies real money for it, a solo founder running the same system has a serious edge.
One more Brett rule for cold outreach: catch people off guard. Memories come from surprise. Buy their product before you pitch them. Send something to their office. A lazy "can I get on a call?" message signals zero effort, and it gets zero replies.
Selling it: say your price, then shut up
Don't sell over text. The message has one job: book the call.
On the call, pitch the outcome in the customer's language, not yours. Brett's example comes from his sneaker bot days: sellers didn't care about the platform he built. They cared that they no longer had to manually reset license keys and chase Venmo requests at 10 p.m. on a Friday.
Then the pricing moment. Say "For this, I charge $2,000." And stop talking.
Most people get nervous and keep explaining. Every extra word discounts you.
If you're brand new and terrified of that silence, Brett's fix is one I love: do your first one or two clients free or heavily discounted. Not forever. Once or twice. You're not buying revenue, you're buying a case study, and one real case study sells the next ten clients better than any pitch you'll ever write.
Delivery: the boring part that decides everything
Two unglamorous details from the course that separate real operators from dabblers:
- Onboarding. The moment someone pays, welcome them and collect everything up front: logins, brand assets, content, business info. Chasing a client for their logo in week three kills your margin and their trust.
- The website. One page, built with Claude Code in an afternoon: what you do, who it's for, the result you get, how it works, case studies, and a calendar link. Brett's agency site took 3 years to build out. Yours needs five sections and a booking button.
You don't need to know everything before you start. You figure it out as you go, and the client is paying you to figure it out.
My take: this is the 2026 version of "learn to code"
A surprising number of the bootstrapped founders I interview started exactly here: sell a service, learn a niche deeply, then productize the thing you keep delivering by hand.
The AI angle just compresses the timeline. Delivery is faster (websites in hours, videos without a crew) and client acquisition is automatable (25 leads a day while you sleep).
What hasn't changed: positioning and sales still decide who wins. The 401(k) guy isn't winning because his AI is smarter. He's winning because his offer is so specific it feels custom-built for the person reading it.
Pick one service. Pick one niche you understand. Get one case study. Everything compounds from there.
FAQ
How much money do you need to start an AI services business?
Almost none. Brett's stack is a Claude subscription, a Calendly account, and a one-page website you build yourself with AI. The real investment is time: daily outreach and your first free case study. Compare that to a SaaS, where you build for months before your first dollar.
Which AI service is best for beginners?
Clipping and UGC pay fastest because the budgets are public and you don't need sales calls (Evan made $17K in 3 months as a college student). AI implementation and app building earn the most per client but need more skill and a real sales process. Start where your existing skills are, not where the ceiling is highest.
How do you get your first client with no portfolio?
Do the work free or discounted for one or two clients in exchange for a written case study and a referral. Then niche your offer down (industry + specific result + specific method) so you're not competing with every freelancer on earth. Specific offers get replies. Generic ones get ignored.
Can an AI services business turn into a SaaS?
That's the best-case path. Brett's agency built apps for clients at $6K/month, took equity, and learned exactly what the market paid for. Service revenue funds you today while client work shows you which tool to productize tomorrow. Plenty of founders on my podcast started as agencies.
Want the founder version of this playbook?
Every week on the Profitable Founder Podcast I sit down with bootstrapped founders making $100K to $10M a year and pull apart exactly how they got their first clients, priced their offers, and productized their way out of service work.
Listen to the latest episode →